NUMI by Agentic AI — Business Plan

1. Executive Summary

NUMI by Agentic AI — Version 1.0 | February 2026 | Confidential

Business Overview

Agentic AI Ltd. is an Israeli technology company developing Numi — a digital baby sleep consultant powered by agentic artificial intelligence (Agentic AI). Numi is not another sleep app and not a chatbot that answers questions. She is an autonomous AI agent that manages the baby's sleep training process end-to-end — diagnosing, building a personalized plan, accompanying parents in real time (even at 2:00 AM), sending reminders, encouraging, identifying problems before they occur, and adapting herself over months and years. All of this in Hebrew, via WhatsApp, at one-tenth the cost of a private sleep consultant.

Numi's professional foundation is the proven methodology of Dorit Kreizer from her book "Shhhh... At Night We Sleep" and years of clinical experience with thousands of families. The company is not starting from scratch — it is scaling an existing, successful business through technology. A proof of concept (POC) is already operational today.

Key Metrics at a Glance

$100K

Funding Required

280

Break-Even Customers

$1.25B

Global Market (2033)

x15–x30

Expected ROI (Base)

Parameter	Value
Legal Name	Agentic AI Ltd.
Product	Numi — Agentic AI Baby Sleep Consultant
Founders	Rotem Levi (CTO) and Dorit Kreizer (Sleep Expert)
Ownership Split	50% / 50%
Stage	Active POC, Pre-revenue
Funding Required	$100,000 (~370,000 ILS) — Owner's Loan
Monthly Burn	~$7,000 (~25,900 ILS)
Zero-Revenue Runway	~14 months
Break-Even Point	280 customers (2.6% of Israel's addressable market)
First Revenue	Month 3–4 (leveraging Dorit's existing audience)
Global Market	$420M (2024) → $1.25B (2033), CAGR 12–13%
Expected Return (Base Case)	x15–x30 on investment

The Problem

37% of parents in Israel report baby sleep problems. 72% of mothers suffer from poor sleep. This is a crisis that impacts mental health, couple relationships, work performance, and quality of life.

All existing solutions fall short:

Existing Solution	Price	Key Limitation
Private sleep consultant	1,500–3,000 ILS	Not available at 2:00 AM; limited to business hours
Books and courses	50–500 ILS	Static, no personalized guidance
Facebook groups	Free	Unprofessional advice, contradictory information
Global apps (Huckleberry, Pampers)	$60–$180/year	English only, reactive (only answer questions), no WhatsApp

The critical gap: No solution on the market — neither in Israel nor globally — provides an autonomous AI agent that accompanies parents in real time, in Hebrew, via WhatsApp.

The Solution: Numi

Numi is a multi-agent system (Multi-Agent System) built on Google ADK and Gemini 2.0 Flash, comprising four specialized agents:

1. Sleep Agent — "Dorit's Clock"

Manages a real-time delay timer, accompanies the parent minute-by-minute during training nights

2. Nutrition Agent

Analyzes correlations between eating and sleeping, identifies patterns

3. Siblings Agent

Coordinates sleep schedules between children, prevents family "meltdowns" (an exclusive advantage — no competitor offers this)

4. Long-Term Memory

Tracks across months and years, identifies regressions, learns across children

The fundamental difference: All competitors — including Huckleberry's Berry, launched in February 2026 — are reactive: the parent asks a question, the AI answers. Numi is agentic: she initiates contact, manages processes, alerts before problems arise, and accompanies in real time. This is the difference between a book you read and a consultant sitting beside you.

The Market

Market	Year	Size (Births/Year)	Key Advantage	Customer Target
Israel	Year 1	~182,000	Zero AI competitors in Hebrew; WhatsApp 98%; Dorit's brand	100–500
United Kingdom	Year 2	~653,000	CPM 47% lower than US; documented gap in NHS services	300–1,500
United States	Year 3	~4,200,000	Massive market; proven product; VC fundraising	800–4,000

Israel is the ideal starting point: a market small enough for rapid dominance, the highest fertility rate in the OECD (2.9 children/family — making the siblings logic a critical advantage), a WhatsApp-first culture, and zero AI competition in Hebrew.

The Team

Founder	Role	Key Contribution	Cost to Company (Year 1)
Rotem Levi	CTO, Co-Founder	Full technology development: AI architecture, Backend, Frontend. Senior engineer at Bigabid (public company).	$0 — works on Numi evenings and weekends
Dorit Kreizer	Sleep Expert, Co-Founder	Proven methodology, book, brand, existing customer base, organic marketing	$0 fixed — 50% revenue share per customer

Both founders are financially independent. Rotem is funded by his employment at Bigabid; Dorit continues to consult independently. Neither depends on Numi for their livelihood — which reduces pressure on capital and enables sound business decisions.

The synergy: Rotem without Dorit = technology without content and audience. Dorit without Rotem = methodology without a technology product. Together = a complete product with built-in Product-Market Fit.

Investment Structure

Parameter	Details
Amount	$100,000 (~370,000 ILS)
Type	Owner's Loan from Dorit to the company
Interest	0%
Priority	Full repayment before any profit distribution, dividends, or founder salaries
Collateral	Lien on all company assets and intellectual property
External Capital	None — zero dilution, full control by both founders
Governance	Expenditure above $1,000 requires joint approval of both founders
Reporting	Full monthly financial report — revenue, expenses, customers, loan balance

Financial Forecast — Three Scenarios

Metric	Conservative	Base	Optimistic
Year 1 Customers	100	300	500
Year 3 Customers	1,200	3,500	6,000
Year 1 Revenue	$30K (~111K ILS)	$90K (~333K ILS)	$162K (~601K ILS)
Year 3 Revenue	$420K (~1.55M ILS)	$1.22M (~4.5M ILS)	$2.25M (~8.3M ILS)
Year 3 EBITDA	+$70K (~259K ILS)	+$725K (~2.7M ILS)	+$1.55M (~5.7M ILS)
Full Loan Repayment	Month 36	Month 30–32	Month 23
ROI on $100K (Year 3)	x4.2	x12.2	x22.5

Revenue Model: Annual subscription paid upfront ($400–$800/year) with a 50/50 split between the consultant and the platform. Company's share: ~$300 per customer/year. No freemium model — paying customers only.

Why This Is a Good Investment — Seven Reasons

1. The downside is limited and predefined

The maximum loss is $100,000 (~370,000 ILS). No additional guarantees, no future obligations, no debts to third parties. The amount is known from day one.

2. Even in total failure — assets are worth money

In the event of failure (probability: ~5%), the intellectual property assets retain a value of $55,000–$120,000. Net loss in the extreme scenario: $0–$45,000 only.

3. Revenue starts early

Dorit comes with customers, a brand, and reputation. First revenues are expected as early as month 3–4 — not month 12 like a typical startup.

4. Low burn = ample time to correct mistakes

$7,000 per month — 4–14x less than a typical startup. 14 months of runway with zero revenue. Enough time to improve, change strategy, or wind down.

5. Low break-even point

280 customers are sufficient — only 2.6% of the addressable market in Israel (10,800 parents willing to pay). An entirely realistic target.

6. Full control — no external investors

Zero dilution. No board of directors. No aggressive growth pressures. Both founders make every decision together.

7. Excellent risk-adjusted return

Weighted expected value: +$1,027,500 (based on detailed probability distribution in Section 10). For every dollar at risk, the expected value is approximately $10 in return.

The Bottom Line

This is a classic asymmetric bet — limited downside, open upside.

The downside: $100,000 (~370,000 ILS). This is the known amount upfront. It does not grow, inflate, or surprise. Even in the worst-case scenario, the company's assets retain value that offsets a significant portion of the loss.

The upside: 50% ownership in an AI platform in a billion-dollar+ market. Annual revenue of $1.2M+ in the base case by Year 3. Estimated company valuation of $3M–$24M (based on 10–20x revenue multiples). A loan repaid with first priority. And all of this — without giving up a single shekel of ownership to an external investor.

In numerical terms: For every $1 at risk, the expected value is $10 to $50 in return.

The amount is protected by seven mechanisms: asset lien, repayment priority, monthly reporting, expenditure governance, low burn rate, existing customer base, and residual value in failure.

We are asking for $100,000. In return — 50% ownership in the company, first-priority loan repayment, equal control in every decision, full transparency, and a technology asset that builds on Dorit's methodology and amplifies her reach 100x.

2. Company Overview and Founders Agreement

NUMI by Agentic AI — Version 1.0 | February 2026 | Confidential

2.1 Company Description

Detail	Information
Legal Name	Agentic AI Ltd.
Trade Name	Numi by Agentic AI
Legal Structure	Limited Company (Ltd.) registered in Israel
Headquarters	Israel
Year Founded	2025
Current Stage	POC / Pre-revenue
Team Size	2 co-founders
Initial Capital	$100,000 (~370,000 ILS) — Owner's Loan

Agentic AI is an Israeli technology company developing agentic AI solutions for the parenting and child health domain. The company's flagship product, Numi, is the first of its kind — an autonomous AI-powered baby sleep consultant. The platform provides personalized, adaptive guidance to parents through an AI agent that learns, adapts, and operates autonomously — available 24/7, in Hebrew, via WhatsApp.

Unlike a typical technology startup, Agentic AI is built on an existing and proven asset: Dorit Kreizer's methodology from her book "Shhhh... At Night We Sleep", years of clinical experience with thousands of families, and an active customer base. The company is not starting from scratch — it is scaling an existing, successful business through technology.

2.2 Mission Statement

Numi's mission is to democratize access to professional baby sleep consulting through AI, making Dorit Kreizer's proven methodology accessible to every parent in Israel and beyond — 24/7, in Hebrew, via WhatsApp. No parent should have to choose between financial well-being and the sleep health of their family.

Value	Practical Meaning
Accessibility	Professional consulting at one-tenth the cost of a private consultant ($600/year vs. $500–$1,500 per session)
Availability	24/7, including nighttime hours when distress peaks — 2:00 AM
Language	Hebrew (launch), English (UK, Year 2)
Channel	WhatsApp — where parents already are
Methodology	Based on a proven method with years of success, not generic internet advice

2.3 Vision Statement

Within 5 years, Numi will become the trusted AI-powered sleep training partner for families in Israel, the UK, and beyond. We envision a world where every new parent has an expert-level sleep consultant in their pocket — AI-powered, learning, adapting, and growing with the family over years and across children.

Time Horizon	Target
Year 1	Active product in Israel with 300 paying customers, revenue covers expenses
Year 2	Expansion to the UK, 1,200 customers, full profitability
Year 3	US market entry, 3,500 customers, $1.2M+ revenue
Year 5	Leading sleep consulting platform in the English-speaking world, B2B partnerships with HMOs and pediatric clinics, multi-language system

2.4 Intellectual Property Foundation

Numi's AI is built on the proprietary methodology from Dorit Kreizer's book — "Shhhh... At Night We Sleep". This book represents years of clinical experience distilled into a structured, proven baby sleep training model. The methodology serves as the core knowledge base for Numi's AI agent, giving it clinical depth and experiential insight that competitors building on generic sleep advice cannot replicate.

IP Asset	Description	Ownership
Sleep Methodology	Dorit's complete method codified into AI decision trees and conversation flows	Licensed to the company
"Dorit's Clock"	A unique graduated response delay system — a core principle of the method	Licensed to the company
Multi-Agent Architecture	Sleep agent, nutrition agent, siblings agent, long-term memory	Owned by the company
RAG Knowledge Base	Vector database from Dorit's book, consulting experience, and professional content	Owned by the company
Source Code	Numi platform, AI agents, interfaces, infrastructure	Owned by the company
Data Repositories	Customer data, usage patterns, AI insights	Owned by the company

Guiding Principle: All data and IP belong to the company (Agentic AI Ltd.) — not to any founder personally. In the event a founder departs, the company retains all intellectual property, code, and data.

2.5 History and Milestones

2024 Q3

Founders identified baby sleep as a critical pain point — initial market research, parent interviews

2024 Q4

Partnership between Rotem and Dorit established — unique combination of technology + clinical expertise

2025 Q1

Market research completed; competitive analysis (11 competitors) — understanding the market, gaps, and opportunities

2025 Q2

Technology architecture designed; Google ADK selected — infrastructure selection, system design

2025 Q3–Q4

POC developed on Google ADK + Gemini — technical proof of concept

2026 Q1

Business plan completed; partnership with Dorit formalized — founders agreement document, terms defined

2026 Q2

MVP development begins with $100K funding — initial investment, build begins

2026 Q3–Q4

First paying customers in Israel — business model validation, first revenues

How the Idea Was Born: The idea for Numi was born at the intersection of two worlds: Rotem's technology expertise in AI and complex systems, and Dorit's clinical experience in sleep consulting for thousands of families. Rotem identified that the agentic AI field was reaching a maturity point that, for the first time, enables taking a proven professional methodology and transforming it into an autonomous agent. Dorit recognized that demand for her services exceeded her physical capacity to deliver them. Numi removes that limitation.

2.6 Founding Team

Rotem Levi — Co-Founder, CTO

BackgroundSenior Software Engineer at Bigabid (private AdTech company)

ExpertiseAI/ML, Systems Architecture, Google ADK, Python, Cloud

EducationSoftware Engineer

CurrentFull-time at Bigabid; Numi evenings/weekends (~15–20 hrs/week)

Salary (Year 1)$0 from Numi

Responsibilities: Full technology development (AI, Backend, Frontend), building and maintaining AI agents, cloud infrastructure, security, strategic and financial decisions (jointly with Dorit).

Transition: Will go full-time when monthly net profit reaches 2x his net salary (~70,000 ILS/month) for 3 consecutive months — a data-driven approach, not a "leap of faith."

Dorit Kreizer — Co-Founder, Sleep Expert

BackgroundLeading baby sleep consultant in Israel; author of "Shhhh... At Night We Sleep"

ExpertiseBaby sleep methodology, parental counseling, content creation, brand building

AudienceThousands of followers; established reputation in the Israeli market

Salary (Year 1)$0 fixed — 50% revenue share per customer payment

Responsibilities: Methodology and content (feeding the AI knowledge base, validating responses), customer relations, marketing and brand, strategic and financial decisions (jointly with Rotem).

The Synergy: Why This Partnership Works

Component	Rotem	Dorit	Together
Technology	AI, architecture, development	—	Advanced technology product
Methodology	—	Years of experience, book, proven method	AI with genuine clinical depth
Audience	—	Thousands of followers, reputation	Customers from day 1, not from zero
Brand	—	"Shhhh... At Night We Sleep"	Recognized brand that builds trust
Risk	Employed at Bigabid, no financial risk	Continues independent consulting	Both founders financially stable
Business Model	Builds the platform	Provides the content + customers	Product + Market = Built-in PMF

2.7 Advisory Board

Agentic AI is actively building an advisory board with expertise in three critical domains:

Pediatric Sleep Medicine

Why: Clinical credibility, regulatory navigation, methodology validation

Profile: Pediatrician specializing in sleep, or certified neonatal nurse

Compensation: 0.25–0.5% equity, quarterly meetings

AI/ML Engineering

Why: Architecture review, model safety, scaling guidelines

Profile: Senior AI engineer from a leading AI company or research lab

Compensation: 0.25–0.5% equity, monthly meetings

SaaS Growth

Why: GTM strategy, B2B sales, unit economics optimization

Profile: Former VP Growth or CMO at a B2C SaaS company with $10M+ ARR

Compensation: 0.25–0.5% equity, monthly meetings

2.8 Founders Agreement / Term Sheet

Important Note: This section constitutes a business summary of the founders agreement terms. The binding agreement will be signed as a separate legal document, with legal counsel, before capital is transferred to the company. All clauses below form the basis for drafting the legal agreement.

A. The Parties

Party	Name	Role	Ownership Percentage
Founder 1	Rotem Levi	CTO, Co-Founder	50%
Founder 2	Dorit Kreizer	Sleep Expert, Co-Founder	50%

B. Capital Structure and Initial Investment

Parameter	Details
Funding Type	Owner's Loan from Dorit to the company
Loan Amount	$100,000 (~370,000 ILS)
Transfer Date	Upon company incorporation and signing of the legal agreement
Interest	0% (interest-free)
Collateral	Lien on company assets and intellectual property (see Section G)
Repayment Priority	Full priority over profit distribution, dividends, and founder salaries
Target Repayment	18–24 months from company revenues
External Venture Capital	None — zero dilution, full control by both founders

Agentic AI Ltd.
    |
    +-- 50% Rotem Levi (Co-Founder, CTO)
    |
    +-- 50% Dorit Kreizer (Co-Founder, Sleep Expert)
    |
    +-- Owner's Loan: $100,000 from Dorit to the company
         (Full repayment priority, before any distribution)

C. Founder Contributions

Domain	Rotem Levi	Dorit Kreizer
Financial Capital	—	$100,000 Owner's Loan
Technology	Full development: AI, Backend, Frontend, Infrastructure	—
Methodology	—	Proven sleep method, book content, clinical experience
Time Invested (Year 1)	~15–20 hours/week (evenings, weekends)	~15–25 hours/week (consulting, content, marketing)
Audience/Customers	—	Existing customer base, thousands of followers, reputation
Marketing and Brand	—	"Shhhh... At Night We Sleep" brand, social media presence
Strategic Decisions	Joint	Joint
Financial Management	Joint	Joint

D. Compensation and Salary Policy

Year 1: Zero Salaries

Founder	Monthly Salary	Income Source	Note
Rotem	0 ILS from Agentic AI	Salary from Bigabid (52,323 ILS gross)	Works on Numi during off-hours
Dorit	0 ILS fixed	Revenue Share — 50% of every customer payment	Continues independent consulting

After Rotem's Transition to Full-Time:

Component	Terms
Rotem's Salary	Market rate: 50,000–55,000 ILS/month gross (equivalent to current salary)
Transition Trigger	Monthly net profit >= 2x Rotem's net salary, for 3 consecutive months
Dorit's Salary	Revenue share; option to transition to fixed salary in Year 2+
Priority Order	Loan repayment > Operating expenses > Founder salaries > Dividends

Guiding Principle: Founder salaries are not paid as long as the loan has not been fully repaid, unless both founders agree otherwise in writing. Dorit's revenue share is not considered "salary" for purposes of this clause — it is part of the company's revenue model.

E. Revenue Sharing Model

Component	Value
Average Price Per Customer	$600/year (~2,220 ILS)
Split	50% to the consultant (Dorit), 50% to the company (Agentic AI)
Company's Share Per Customer	$300/year = $25/month
Payment Model	Annual, upfront

Customer pays $600/year
         |
    +----+----+
    |         |
   50%       50%
    |         |
  $300      $300
 To the    To the
consultant  company
 (Dorit)   (Agentic AI)

F. Cash Flow Distribution Priority

Priority 1:  Ongoing operating costs (~$7,000/month = ~25,900 ILS)
             Servers, APIs, marketing, legal, accounting
             |
             v
Priority 2:  Owner's Loan repayment ($100,000)
             Every surplus dollar is directed to repayment until fully paid
             |
             v
Priority 3:  UK market expansion fund
             Allocation for international growth
             |
             v
Priority 4:  Founder salaries (if applicable)
             Only after full loan repayment
             |
             v
Priority 5:  Profit distribution (dividends) by ownership percentage
             50/50 -- only after all of the above

Practical meaning: Dorit receives her funds in two ways — (1) ongoing revenue share (50% per customer, from day one), and (2) loan repayment from company surpluses (with first priority). Only after the loan has been fully repaid will additional dividends be considered.

G. Lender Protections (Dorit's Investment Protection)

The loan is secured by seven built-in protection mechanisms:

Protection 1: Company Asset Lien

The loan is secured by a lien on all company assets, including: the Numi platform (source code, architecture, design), data and customer databases, AI models, algorithms, system content, and rights to use the methodology. Even in total failure, the IP assets retain value estimated at $55,000–$120,000.

Protection 2: Liquidation Priority

In the event of company liquidation, priority order: (1) Payment of external debts, (2) Owner's Loan repayment — $100,000 (full priority), (3) Distribution of remaining assets by ownership percentage (50/50). Dorit is "first in line" before any asset distribution to shareholders.

Protection 3: Monthly Financial Reporting

Monthly: Profit & Loss, Cash Flow, Customer Status, Loan Balance. Quarterly: Strategic Review. Annually: External CPA audit (when revenue justifies). Deviations reported as needed with corrective actions.

Protection 4: Expenditure Governance

Up to $1,000: either founder independently. Above $1,000: joint approval required. New contracts, hiring, strategic decisions, pricing changes, budget overruns — all require joint approval.

Protection 5: Low Burn Rate

Monthly expenses: ~$7,000 (vs. $30K–$100K typical startup). Zero-revenue runway: ~14 months (vs. 6–12 months). Daily expense: ~$233. Numi's burn rate is 4–14x lower than a typical startup.

Protection 6: Existing Customer Base

Dorit has an active customer base, a popular book, and social media channels with thousands of followers. The probability of zero revenue over an extended period is extremely low. Revenue is expected as early as month 3–4.

Protection 7: Residual Value in Failure

Even in the worst-case scenario (total failure), the assets built retain value:

Asset	Estimated Value
Source code and platform	$30,000–$60,000
Codified methodology	$10,000–$20,000
Intellectual property rights	$10,000–$30,000
Data repository and insights	$5,000–$10,000
Total	$55,000–$120,000

H. Data and Intellectual Property Ownership

Asset	Ownership	Details
Customer Data	The Company	All user data, usage patterns, insights
Source Code	The Company	All code written for Numi stored in a company-owned repository
AI Models	The Company	Models, fine-tuning, embeddings, agent configurations
RAG Knowledge Base	The Company	Vector database, processed content, training data
Original Methodology	Dorit (licensed)	Dorit licenses the company to use it; see Section I
"Numi" Brand	The Company	Trademark, logo, design, marketing materials
"Shhhh..." Brand	Dorit	Book title and personal brand remain Dorit's

Guiding Principle: If a founder leaves the company — all data, code, models, and the knowledge base remain with the company. No founder is permitted to take any company data or IP with them.

I. Intellectual Property Licensing Arrangement

Parameter	Terms
Licensor	Dorit Kreizer
Licensee	Agentic AI Ltd.
Subject of License	Sleep methodology from the book "Shhhh... At Night We Sleep"
License Type	Exclusive for use in a digital platform (AI agents)
Scope	Baby sleep consulting via AI
Duration	As long as the company is active
Consideration	50% company ownership + 50% revenue share
Restrictions	Does not prevent Dorit from continuing personal consulting, selling books, or giving lectures
New IP	Any new IP created for Numi belongs to the company
Dissolution	If the company dissolves — the methodology license reverts to Dorit

J. Vesting (Share Vesting)

Parameter	Terms
Vesting Schedule	4 years with a 1-year Cliff
Cliff	12 months — if a founder leaves before one year, they forfeit all shares
Post-Cliff Vesting	Monthly vesting — 1/48 of shares each month
Accelerated Vesting	In the event of an Exit (sale/IPO) — full and immediate vesting for both founders

Illustrative Example:

Period	Shares Vested (per founder)	Percentage of 50%
Month 0–11	0% (Cliff period)	0%
Month 12	12.5% (25% of 50%)	25%
Month 24	25% (50% of 50%)	50%
Month 36	37.5% (75% of 50%)	75%
Month 48	50% (100% of 50%)	100%

K. Decision Making and Dispute Resolution

Decision Type	Who Decides	Examples
Routine Operational	Each founder independently	Purchasing dev tools, paying recurring bills (up to $1,000)
Material Decision	Both founders jointly	Expenditure above $1,000, hiring, B2B partnerships, pricing changes
Strategic	Both founders jointly	Fundraising, selling the company, entering a new market, changing the business model

Deadlock Resolution Mechanism:

Step 1: Discussion between founders (up to 14 days)
        |
        v (if no agreement reached)
Step 2: Mediation -- referral to an agreed-upon external professional (up to 30 days)
        |
        v (if mediation fails)
Step 3: Binding arbitration -- arbitrator predetermined in the agreement

L. Exit Provisions

If a founder leaves the company:

Condition	Details
IP and Data	Remain with the company — no exceptions
Vested Shares	The departing founder retains vested shares (per vesting schedule)
Unvested Shares	Revert to the company
Right of First Refusal	The remaining founder has ROFR to purchase the departing founder's shares
Valuation	Share price determined by an independent external appraiser
Non-Compete Period	12 months — no directly competing product
Confidentiality	Ongoing confidentiality obligation without time limit

Scenario: Dorit Leaves

Detail	What Happens
Methodology codified in AI	Remains with the company (company IP)
Book "Shhhh..."	Returns to Dorit (her personal brand)
Methodology license	Rotem needs alternative methodology or re-acquire license
Vested shares	Dorit retains them
Owner's Loan	Continues to be repaid per loan terms

Scenario: Rotem Leaves

Detail	What Happens
Source code	Remains with the company (company IP)
Architecture and AI models	Remain with the company
CTO replacement	Dorit will need to recruit a replacement CTO
Vested shares	Rotem retains them

M. Dissolution

Priority Order	Details
1. External Debts	Payment of all obligations to third parties
2. Owner's Loan Repayment	$100,000 to Dorit (full priority)
3. Remaining Assets	Distribution by ownership percentage: 50/50
4. Methodology License	Reverts to Dorit in full
5. Source Code	Both founders may use relevant portions (per agreement)

N. Special Conditions

Condition	Details
Anti-Dilution	No external fundraising without consent of both founders
Share Sale Restriction	No share sale to a third party without first offering to the other founder (ROFR)
Insurance	D&O insurance when revenue justifies
Incorporation	Company incorporation in Israel before capital transfer
Bank Account	Bank account in the company's name only
Accounting	External CPA from day one
Agreement Updates	May be updated with written consent of both parties

2.9 Agreement Summary — Consolidated Table

Clause	Terms
Ownership	50% Rotem / 50% Dorit
Investment	$100,000 Owner's Loan from Dorit
Interest on Loan	0%
Repayment Priority	Loan > Operations > Founder Salaries > Dividends
Target Repayment	18–24 months
Collateral	Lien on company assets + IP
Founder Salaries (Year 1)	$0 for both
Revenue Share	50% consultant / 50% company
Vesting	4 years, 1-year Cliff
IP Ownership	The company (except Dorit's personal brand)
Methodology License	Exclusive to the company; reverts to Dorit upon dissolution
Data Ownership	100% the company
Code Ownership	100% the company
Operational Decisions	Up to $1,000: independent; above $1,000: joint
Deadlock	Discussion > Mediation > Binding Arbitration
Founder Departure	IP stays with the company; ROFR on shares; vested shares retained
Dissolution	Loan first > remainder 50/50; methodology reverts to Dorit
Reporting	Monthly (financial) + Quarterly (strategic)
Governance	Expenditure above $1,000 requires joint approval
External Capital	Only with consent of both founders

2.10 Why This Structure Is Fair — For Both Parties

From Dorit's Perspective

Question	Answer
What protects my money?	Full repayment priority, asset lien, monthly reporting, expenditure governance
What if the project fails?	Max loss $100K; IP assets worth $55K–$120K; methodology reverts to me
What if Rotem leaves?	IP stays with the company; I hold 50%; I recruit a replacement CTO
What is my upside?	50% of every customer (from day 1) + 50% company ownership + loan repayment
Who is in control?	Equal control — no material decision without my consent

From Rotem's Perspective

Question	Answer
Why 50% and not more?	Dorit brings methodology, audience, brand, and $100K. Fair value
What if Dorit leaves?	IP codified in AI stays; I need alternative methodology
When do I start earning?	Full salary when company proves capacity (2x salary, 3 months)
Who controls technology?	I do — full CTO, my technology decisions
What protects me?	Vesting, ROFR, IP belongs to company, non-compete

Why 50/50 and Not a Different Split?

Argument	Explanation
Rotem brings the technology	Full development of an agentic AI platform — market value $100K–$200K+
Dorit brings the capital	$100K — the money that enables the company to launch
Dorit brings the methodology	Years of clinical experience, book, proven method — irreplaceable
Dorit brings the audience	Customers from day 1 — the most expensive thing in a startup
Both dedicate significant time	15–25 hours/week each in Year 1
Both are essential	No product without technology; no product without methodology

A 50/50 partnership reflects the fact that both founders are equally essential to Numi's success. Rotem without Dorit = technology without content and audience. Dorit without Rotem = methodology without a product. Together = a complete product with built-in Product-Market Fit.

2.11 Execution Timeline

March 2026 — Step 1

Sign founders agreement (legal, with attorney) — Both founders

March 2026 — Step 2

Incorporate Ltd. company in Israel — Rotem + CPA

March 2026 — Step 3

Open bank account in company name — Both founders

April 2026 — Step 4

Transfer Owner's Loan ($100K) to company account — Dorit

April 2026 — Step 5

Begin MVP development — Rotem

June–July 2026 — Step 6

MVP ready; beta launch — Both founders

Aug–Sep 2026 — Step 7

First paying customers — Dorit (referrals) + Marketing

December 2026 — Step 8

Year 1 summary; annual financial report; Year 2 planning — Both founders

3. Product and Technology

Numi's Multi-Agent AI Architecture and Product Layers

3.1 Product Overview

Numi is a digital sleep consultant powered by artificial intelligence, operating via WhatsApp and a web interface. Unlike existing sleep apps on the market that provide generic tips or passive tracking, Numi is an autonomous agent — a system that doesn't just answer questions, but initiates action, monitors in real time, reminds, accompanies, and predicts.

Professional Foundation

All of Numi's knowledge is built on Dorit Kreizer's methodology from her book "Shhhh... At Night We Sleep" and her cumulative coaching experience — not from generic internet sources, but from a specific, proven method.

Core Technology: RAG

Numi is built on a RAG (Retrieval Augmented Generation) architecture — the AI retrieves the most relevant information from Dorit's knowledge base and constructs a personalized response. Like asking Dorit herself — except she's available 24/7.

Current Status: Active POC

A proof of concept is already operational on Google ADK and Gemini 2.0 Flash infrastructure, demonstrating professional, personalized conversations with parents.

The Fundamental Difference: AGENTIC vs. REACTIVE

This is the critical difference between Numi and every competitor on the market, including Huckleberry:

Feature	Competitors (REACTIVE)	Numi (AGENTIC)
When it operates	Only when the parent asks a question	Initiates contact on its own, even when the parent doesn't reach out
Type of assistance	Answer to a single question	Continuous accompaniment — reminders, encouragement, alerts
Process management	The parent manages the process themselves	Numi manages the sleep training process
Problem detection	Responds to the parent's report	Identifies patterns and alerts before a problem occurs
Timer/Clock	None	Manages "Dorit's Clock" in real time
Cross-analysis	Analyzes sleep in isolation	Connects nutrition, sleep, development, and siblings

Simply put: Other apps are like a book you can open and read. Numi is like a consultant sitting beside you, watching the baby, and telling you what to do — before you even knew a question existed.

3.2 Multi-Agent System Architecture

Why Multi-Agent? The Core Idea

Numi is not a chat app. It is a system of intelligent agents that learn from one another.

"Regular" artificial intelligence — like ChatGPT or a chat bot — is a single system trying to do everything. That works for general questions, but when you need to manage a complex process like baby sleep training — with dozens of variables changing every day — a single system is insufficient.

Numi's solution is a multi-agent architecture: instead of one bot trying to know everything, there are four specialist agents, each with a clear role, working together and sharing information among themselves. Like a medical team where each specialist brings their expertise — but they all communicate and coordinate care.

Numi's Four Agents

+------------------------------------------------------------------+
|                        Numi Brain                                 |
|                (Central coordination between agents)              |
|                                                                   |
|  +--------------+  +--------------+  +--------------+             |
|  | Sleep Agent  |  |  Nutrition   |  |  Siblings    |             |
|  |              |  |    Agent     |  |    Agent     |             |
|  |              |  |              |  |              |             |
|  | "Dorit's     |  | Correlation  |  | Schedule     |             |
|  |  Clock"      |  | between food |  | coordination |             |
|  | Timer mgmt   |  | and sleep    |  | between      |             |
|  | Real-time    |  | Pattern      |  | children     |             |
|  | training     |  | detection    |  | Preventing   |             |
|  |              |  | Nutritional  |  | "meltdowns"  |             |
|  |              |  | guidance     |  |              |             |
|  +--------------+  +--------------+  +--------------+             |
|                          |                                        |
|              +-----------------------+                            |
|              |  Long-Term Memory     |                            |
|              |                       |                            |
|              | Tracking over years   |                            |
|              | Predicting regressions|                            |
|              | Cross-child learning  |                            |
|              | Family profile        |                            |
|              +-----------------------+                            |
+------------------------------------------------------------------+
          |                    |                    |
    +----------+        +-----------+       +------------+
    | WhatsApp |        |    Web    |       | Social     |
    | Business |        | Dashboard |       | (FB/IG)    |
    |   API    |        |           |       |            |
    +----------+        +-----------+       +------------+

Agent 1: Sleep Agent — "Dorit's Clock"

Role: Managing the sleep training process in real time, including managing Dorit's unique delay timer.

What is "Dorit's Clock"? In Dorit Kreizer's methodology, sleep training involves a process of graduated delays — the parent waits a defined period before entering the baby's room. The times increase gradually: 3 minutes, 5 minutes, 8 minutes, and so on. Most parents struggle to manage this on their own at 2:00 AM.

What the agent does:

Manages the timer in real time via WhatsApp — "3 minutes have passed. Don't go in yet. You're doing great. 2 more minutes."
Adjusts delays based on baby's age, history, and specific night in the process
Provides real-time encouragement — "It's normal that he's crying. You're on night two."
Identifies unusual situations — abnormal crying, fever, vomiting — and recommends checking
Documents every night — time to fall asleep, wake-ups, crying durations

Real example: 02:14 AM. The baby wakes up and cries. Mom sends a WhatsApp message: "He woke up again." Numi responds within seconds: "Night two, second wake-up. According to the plan — don't go in now. Starting a 5-minute countdown. I'm with you." Then, every minute, an encouragement message. After 5 minutes: "Time's up. Go in, place your hand down, 30 seconds, and leave."

Why this matters: No competitor does this. No other app manages a real-time timer and accompanies the parent at the most difficult moment. This is the guidance that until now only a human sleep consultant could provide — and it costs $500–$1,500.

Agent 2: Nutrition Agent — Correlation Between Food and Sleep

Role: Analyzing causal connections between what the baby eats and when, and the quality of their sleep.

Why a separate agent? Many parents are unaware that nutrition directly affects sleep. A baby who ate too large a meal close to bedtime may wake from reflux. A baby who didn't eat enough will wake from hunger. The relationship is complex and changes with age.

What the agent does:

Receives feeding/nursing information through simple WhatsApp reports
Correlates eating data with sleep data — analyzes patterns over weeks
Identifies and reports correlations: "On days when he eats dinner before 6 PM, sleep is 40% better"
Coordinates with the sleep agent — if the baby ate little, the sleep agent expects a hunger wake-up
Guides through weaning processes — nighttime nursing weaning, transition to solids

The value: No human sleep consultant, no matter how good, can analyze hundreds of data points and find statistical correlations between nutrition and sleep. This agent gives Numi an advantage that simply does not exist in human consulting.

Agent 3: Siblings Agent — The Problem No Competitor Solves

Role: Coordinating sleep and nap schedules between two or more children in the same household.

Why is this the hardest problem? Parents of a baby + a toddler face exponential complexity. Example: the 6-month-old needs a nap at 3:00 PM. The 3-year-old returns from daycare at 3:15 PM. If the baby doesn't fall asleep before the older sibling arrives — noise and disruption cause a "meltdown" that ruins both children's sleep.

What the agent does:

Manages a unified schedule for all children — sleep times, naps, meals, daycare
Identifies conflicts before they happen: "The nap at 3 PM overlaps with the older one's return. Move to 2:30 PM."
Coordinates developmental transitions — when a baby goes from 3 naps to 2
Manages complex nights — considers noise impact on siblings
Prevents family "meltdowns" — precise timing for all variables

Decisive competitive advantage: We surveyed 11 competitors (Huckleberry, Smart Sleep Coach by Pampers, Nanit, and more). None of them offers a solution for coordinating sleep between siblings. All focus on a single child only. 60% of families with a sleep-problem baby also have additional children. This is a massive, unserved market.

Agent 4: Long-Term Memory — The Agent That Remembers Everything

Role: Maintaining a complete history of every child over years, identifying developmental regressions before they occur, and learning across children.

What the agent does:

Accumulates knowledge over time — months and years of data, not just a week
Identifies regressions before they occur: "4 months old. A sleep regression is expected in 2–3 weeks."
Identifies developmental transitions — from 2 naps to 1, crib to toddler bed, diaper weaning
Cross-child learning: "With the older sibling, method X worked at 4 months. Let's try that here too."
Builds a family profile — learns the family's sleep DNA and improves with each child

Real example: The Cohen family used Numi with their first child at 5 months. Two years later, a second child is born. Numi says: "Welcome back! With Yoav, the graduated delay method worked excellently in 4 nights. His temperament was similar. Let's start with the same structure, with adjustments for her age." This is knowledge that no other app on the market can provide.

How the Agents Work Together — A Real-Life Scenario

21:30 — Bedtime

Sleep Agent "Time to start the bedtime routine. Today was a long day — let's start 15 minutes early."

21:32

Nutrition Agent "The last meal was at 7:00 PM, 180 ml. Sufficient — no hunger wake-up expected."

21:33

Siblings Agent "The older one is already asleep. His door is closed. You can proceed with the plan."

21:45

Sleep Agent Parent reports: "I put him down, he's crying." — "Great that you put him down awake. Night three — 5-minute delay. Starting countdown."

21:50

Sleep Agent "5 minutes have passed. He's still crying — but notice the intensity has decreased. Go in for 30 seconds, place your hand, and leave."

21:55

Sleep Agent Parent: "He fell asleep!" — "Amazing. Night 3, fell asleep in 10 minutes. 60% improvement from night 1. See you tomorrow."

22:00

Long-Term Memory Records: time to fall asleep, number of delays, crying intensity, and the night's success. This information will be part of the child's history — for years to come.

3.3 Product Evolution: From "Emergency" to "Lifelong Accompaniment"

Redefining the annual subscription: not "sleep training" but developmental information management. The parent isn't paying to "fix a problem" — they're paying for a system that knows their child over years and guides them at every stage.

Stage 1: Numi Rescue — "Emergency Week" (Week 1)

Immediate Relief

The parent arrives because the baby isn't sleeping. It's urgent.

Day	Action
Day 0 — Intake	Comprehensive questionnaire: age, weight, health, sleep history, eating habits, family structure. Complete profile.
Day 1 — Plan	Based on Dorit's methodology and intake data: personalized sleep plan, schedule, training method, delay times.
Night 1–3	"Dorit's Clock" active. Numi manages timers, sends encouragement, documents every wake-up, guides minute-by-minute.
Day 4–7	Transition from accompaniment to monitoring: checking habits establish, identifying issues, targeted encouragement.

Success metric: 80%+ of parents see significant improvement within 5–7 nights.

Stage 2: Numi Growth — Ongoing Guidance (Months 1–24)

Proactive Monitoring

After the emergency week, Numi transitions to "observation" mode — watches, learns, intervenes only when needed.

Detects nap transitions (2 naps to 1) based on patterns
Detects regression signs and prepares parents
Suggests schedule adjustments before chaos begins
Weekly reports: "6 nights continuous sleep, one wake-up. Improving."

Value: Prevents problems before they occur. This justifies the annual subscription.

Stage 3: Numi Siblings — Multi-Child Complexity

Premium Feature

A second child is born. Suddenly, everything that worked needs updating.

Synchronizes newborn's schedule with existing toddler's
Prevents family "meltdowns" with precise timing
Coordinates nighttime sleep across siblings
Tailored plan for each child's age and needs

Why premium: No competitor solves this. This is Numi's flagship differentiator for the long term.

Stage 4: Family Memory — Across Years and Children

Retention Moat

The system accumulates knowledge about the family over years and uses it to improve.

Preserves all data from the first child
Uses previous child data to improve recommendations for the next
Builds a "family sleep profile" — genetic sensitivities, method preferences
Anonymous cross-family learning from thousands of families

The real value: The longer they use it, the greater the value. The accumulated data and learning are irreplaceable. This is a retention moat — the harder it is to leave.

3.4 Three Business Product Layers

Beyond the technical architecture, Numi is organized into three business layers that create a complete revenue engine:

Super Numi — Lead Engine

New Users

An AI chat on the consultant's website. At 2:00 AM, "Why won't he sleep?" gets an instant answer using Dorit's method. Simple problems solved directly. Complex cases get booked with Dorit automatically.

Value: Converts cold traffic into warm, paying leads — 24/7, while the consultant sleeps.

Digital Numi — After-Care Guide

Existing Clients

Every paying client gets a "Digital Numi" companion. Accompanies them through the first 3 nights, sends reminders ("Put him down awake"), encourages, answers follow-up questions.

Value: Increases success rates, reduces support time, justifies premium packages.

Omnichannel Agent — The Brain

Digital Presence

Connected to Facebook, Instagram, and WhatsApp. Has read Dorit's book, knows her tone. Answers questions about teething, naps, pacifiers automatically across all channels.

Value: Saves hours of manual responses, ensures every parent gets the right answer immediately.

3.5 Tech Stack

Layer	Technology	Why We Chose It
Agent Orchestration	Google ADK (Agent Development Kit)	Leading platform for multi-agent systems; built-in agent coordination, memory management, tool invocation
Artificial Intelligence	Gemini 2.0 Flash	Fast, cost-efficient, accurate; response time under 2 seconds; low cost enabling scale
Knowledge Base (RAG)	Gemini Embeddings + Vector DB	Dorit's book, method, and case files encoded as embeddings for fast semantic search
Primary Channel	WhatsApp Business API	Where parents already are; no new app download needed; 98% open rates
Visual Interface	Web Dashboard (Next.js)	Dashboard for data, charts, and progress tracking
Social Channels	Facebook Messenger API, Instagram API	Connection to Product 3 (the omnichannel agent)
Cloud Infrastructure	Google Cloud Platform	Reliable, scalable, integrated with Google ADK
Security	AES-256 encryption, TLS 1.3, HIPAA-ready	Full protection of sensitive baby and family data

Existing POC

The proof of concept is currently operational and demonstrates:

Personalized Conversations

RAG Capability

<3s

Response Time (target: <2s)

WhatsApp Operation

3.6 Why This Architecture Changes Everything

Five reasons why Numi's multi-agent architecture is a competitive advantage that is extremely difficult to replicate:

1. Specialization = Higher Quality

Each agent is an expert in its domain. The sleep agent doesn't need to "know" nutrition — it receives information from the nutrition agent. Like a doctor referring to a specialist.

In practice: Huckleberry uses a general AI that tries to cover everything. Numi uses 4 focused agents. The difference in answer quality is significant.

2. Inter-Agent Information Sharing = Unique Insights

Nutrition affects sleep. Sleep affects development. Development affects siblings. Numi's agents share information in real time, generating insights impossible from isolated apps.

In practice: No competitor automatically links eating data with sleep data.

3. Long-Term Memory = Value That Grows (Retention Moat)

The longer time passes, the more Numi knows about the child. This information becomes irreplaceable. Switching loses all history.

In practice: Expected churn rate: less than 5%/month, vs. industry average of 8–12%.

4. Siblings Logic = Absolute Exclusivity

No competitor offers sleep coordination between siblings. Adding this requires rebuilding from the ground up.

In practice: 60% of families with sleep problems have more than one child. Massive unserved market.

5. Proactive vs. Reactive = Happier Parents

Numi acts before a problem occurs. Instead of waiting for the parent to report "he's not sleeping," Numi sends a message: "Regression expected in two weeks. Here's what to do." A parent who receives an alert before a problem is much happier than a parent who has to ask for help after the situation has deteriorated.

3.7 Summary: Why Numi Is Different From Everything on the Market

Criterion	Competitors	Numi
Architecture	Single chatbot or rule-based system	4 specialized agents + central coordination
AI Type	Reactive — answers questions	Agentic — initiates, monitors, alerts, manages
Knowledge Base	Generic / based on general research	Specific, proven methodology (Dorit Kreizer)
Process Management	The parent manages	Numi manages (including real-time timers)
Sibling Coordination	Does not exist	Dedicated agent for multi-child management
Memory	Resets each session / short period	Memory spanning years + cross-child learning
Channels	App only	WhatsApp + Web + Facebook + Instagram
Value Over Time	Decreasing (after problem is solved)	Increasing (more data = better recommendations)

The bottom line: Numi is not another sleep app. It is a complete artificial intelligence system that manages the family's children's sleep — from the first night to years ahead, from the first child to the third, from WhatsApp to Instagram. This is a product that does not exist on the market today, and the multi-agent architecture is the reason it cannot be built without dedicated ground-up planning.

4. Industry & Market Analysis

4.1 Industry Overview

The baby sleep consulting market sits at the intersection of digital health and AI-driven consumer applications. Parents are increasingly turning to technology for solutions previously available only through expensive consultations or generic books.

Metric	Value
Baby Sleep Apps Market (2024)	$420M
Baby Sleep Apps Market (2033)	$1.25B
CAGR (2024–2033)	12–13%
Broader Sleep Products (2023)	$6.5B
Broader Sleep Products (2032)	$10.2B

Baby Sleep Apps Market Growth

Key Industry Players: Huckleberry (VC-backed, AI + expert hybrid), Smart Sleep Coach by Pampers (P&G), Nanit ($70M+), Cradlewise ($27M+), Taking Cara Babies (bootstrapped).

The market is fragmented with no dominant player. Most solutions fall into: tracking apps, expensive hardware, one-time courses, or traditional consulting. None offer a fully autonomous Agentic AI experience in Hebrew.

4.2 Market Strategy: Israel → UK → US

Numi adopts a staged entry strategy starting in the domestic market, reducing risk and leveraging unique competitive advantages at each stage.

Year 1: Israel

Home Market

182K births/year
Hebrew (exclusive)
0 AI competitors
WhatsApp 98% penetration
Dorit's established brand
Goal: 100-500 customers

Year 2: UK

First Expansion

653K births/year
English expansion
CPM 47% lower than US
NHS service gap
Proven product + revenue
Goal: 400-2,000 customers

Year 3: US

Scale Market

4.2M births/year
Massive market
VC-backed expansion
2 years of refinement
Clinical data + success stories
Goal: 1,200-6,000 customers

Why Israel First

Advantage	Details
Zero AI competitors in Hebrew	No direct competitor offers Agentic AI sleep consulting in Hebrew. Market completely open
Dorit Kreizer's brand	Book sold via Kinneret Zmora publisher, Steimatzky bookstores, largest baby retail chain. Immediate credibility
WhatsApp as primary channel	98% penetration—parents already use it 24/7. No app download friction
Tech-savvy parents	"Startup Nation" with 91% smartphone penetration, early AI adoption
Highest fertility in OECD	2.9 children/family = recurring demand across multiple children
Small market = rapid dominance	182K births allows establishing dominance before global expansion

4.3 TAM / SAM / SOM

$1.71BTAM (3 Markets)

$353MSAM

$720K-$3.6MSOM (Y3)

TAM—$1.71B
Israel: $40M, UK: $157M, US: $1.51B (births × sleep problem % × avg spend)

SAM—$353M
Reachable at each stage: Y1 Israel $12M, Y2 +UK $51M, Y3 +US $353M

SOM—$720K-$3.6M (Y3)
Conservative: 1,200 customers × $600 ARPU. Optimistic: 6,000 × $600

Market	Births/Year	Sleep Problems %	Avg Spend	TAM
Israel	182,000	37%	$600	$40M
UK	653,000	40%	$600	$157M
US	4,200,000	60%	$600	$1.51B
Total	5.04M	—	—	$1.71B

4.4 Market Trends & Demand Drivers

1. Digital Health Adoption

Telehealth surged post-COVID. Israel's digital health market projected to reach $651M by 2029 (9.22% CAGR). Parents expect 24/7 mobile access to healthcare guidance.

2. AI Acceptance Rising

77% of consumers open to AI health guidance (Accenture 2024). Agentic AI—where AI takes autonomous action—is the next wave. Numi leads this shift.

3. Parental Mental Health

72% of Israeli mothers suffer poor sleep quality. Postpartum depression affects 1 in 7 globally. Sleep deprivation is a primary contributor. Employers and health funds investing in solutions.

4. High Search Demand

"How to get baby to sleep": 1.2M monthly global searches. Hebrew searches substantial (data being gathered). High-intent parents actively seeking solutions.

5. Agentic AI Mainstream

2025-2026 marks the inflection point. Major tech companies shipping AI agents. Parents expect AI that acts: adjusts plans, sends reminders, coaches in real time.

4.5 Israeli Market Advantages—In Depth

#	Advantage	Data/Explanation
1	Zero AI competitors in Hebrew	All 12 analyzed competitors are English-only. Numi has no direct competition
2	WhatsApp infrastructure	98% adult penetration (ISOC-IL 2024). Parents already live on WhatsApp. No app download friction
3	Dorit's established brand	Published by Kinneret Zmora, sold at Steimatzky, 18 years midwife/psychotherapist, Tel Aviv University lecturer
4	Small market = rapid dominance	182K births/year vs 4.2M in US. Can establish presence quickly before global players notice
5	Early tech adopters	"Startup Nation": 91% smartphone penetration, high openness to AI solutions
6	Highest fertility in OECD	2.9 children/woman. Every family is potential multi-child customer (first, second, third baby)
7	Dense word-of-mouth	WhatsApp groups, Facebook parenting communities, forums—recommendations spread rapidly

4.6 Unmet Needs in Israeli Market

Need	Current State	Numi's Solution
Expert sleep consulting 24/7 in Hebrew	Does not exist. Consultants work business hours	AI agent available anytime via WhatsApp
Ongoing support at reasonable cost	ILS 1,400-3,000 one-time, limited follow-up	Annual subscription ILS 1,500-3,000 with 365 days support
Personalization for specific baby	Generic Facebook advice, books	AI learns each baby's patterns, adapts in real time
Sibling sleep coordination	No structured solution	AI coordinates sibling schedules (critical for 2.9 children/family)
Immediate crisis response	Message consultant (reply morning) or Google at 2 AM	Response in seconds, any hour, based on methodology

4.7 Market Forecasts

Year	Market Size (Global Apps)	Growth	Key Drivers
2024	$420M	—	Baseline
2025	$470M	+12%	AI adoption, insurance pilots
2026	$525M	+12%	Telemedicine integration, clinical studies
2027	$590M	+12%	Global expansion, insurance coverage
2028	$660M	+12%	Agentic AI mainstream, B2B partnerships
2033	$1.25B	+12% CAGR	Market maturity, category leaders

5. Competitive Analysis

5.1 Competitive Landscape Overview

We analyzed 12 competitors across 5 categories. The analysis reveals a fragmented market with no player providing a fully agentic AI experience. In the Israeli market (Numi's Year 1 target), there is currently no direct competitor offering AI-based sleep consulting in Hebrew.

5.2 Critical Update: Huckleberry's Berry (Feb 5, 2026)

Most Important Competitive Development in 2026: Huckleberry launched Berry—a context-aware AI chat designed to reduce mental load for parents. This is the most significant competitive shift in the market.

What Is Berry?

Berry is a conversational AI assistant available 24/7 as part of Huckleberry Premium. Provides guidance on childcare, sleep, feeding, behavior, development.

Feature	Details
Data repository	5 billion data points across Huckleberry platform
AI approach	Bridging "Context Gap"—combining logged child data with pediatric expertise
Expert team	Amber LoRe (Director of Pediatric Expertise) + sleep/lactation/OT experts
Privacy	Does not use personal data to train public models
Pricing	$14.99/month or $119.99/year

Berry's Weaknesses—Numi's Advantages

Factor	Berry (Huckleberry)	Numi
AI Type	Reactive (Q&A chatbot)	Agentic (autonomous, proactive)
Hebrew Support	English only	Native Hebrew from ground up
WhatsApp Integration	App only	Native WhatsApp (98% penetration)
Methodology	Multi-expert, generic	Dorit's method, specific & proven
Sibling Logic	No coordination	Yes (2.9 children/family Israel)
Long-term Memory	No cross-child memory	Longitudinal AI across years/children
Pricing Model	Monthly (high churn)	Annual (high commitment)

The Core Difference: Berry answers questions. Numi manages the sleep training process. Example: Berry waits for parent to ask "Why waking at 3 AM?" Numi sends proactive message at 7 PM: "Based on today's feeding, move bedtime 15 min earlier to prevent the 3 AM waking."

5.3 Direct Competitors—AI Apps

Smart Sleep Coach by Pampers

App

Pricing: Free + $7.99/mo or $59.99/yr | Backing: P&G | AI: Schedule optimization
Strengths: Massive brand, global distribution, P&G resources
Weaknesses: Schedule-based only (not conversational/agentic), slow corporate iteration, no Hebrew/WhatsApp

Huckleberry (including Berry)

App

Pricing: Free + $14.99/mo or $119.99/yr | Backing: VC-backed | AI: Berry (reactive) + expert hybrid
Strengths: Market leader, 4.7/5 reviews (50K+), 5B data points, expert team
Weaknesses: Berry still reactive (not agentic), English only, no WhatsApp, no sibling logic

Baby Needs to Sleep AI

App

Pricing: $4.99/mo or $29.99/yr | Backing: Small team
Strengths: Low price, clear AI positioning | Weaknesses: Limited resources, small base, no unique methodology, no Hebrew

Baby SleepBot AI

App

Pricing: ~$30-50 | Market: Australia
Strengths: Proven consultant methodology (similar to Numi approach) | Weaknesses: Limited to Australia, single consultant dependency

Other AI Apps: Lizzy Sleep (24/7 chat, unclear methodology), Baby Snooze (tracker + AI coaching, mobile-only)

5.4 Indirect Competitors

Hardware

Cradlewise & Nanit

Cradlewise: $1,999 crib + $99/yr, $27M+ Series B. AI bounce, sleep tracking.
Nanit: $299-399 camera + $149/yr, $70M+ funding. Computer vision tracking, breathing monitoring. HQ in Tel Aviv + NYC.
Implication: Hardware validates demand but serves monitoring, not coaching. Numi complements them, can integrate via API.

Course

Taking Cara Babies

Pricing: $79-179 per course (one-time) | Est Revenue: $10M+ annually
Strengths: Massive brand, neonatal nurse, proven content
Weaknesses: One-time purchase, no personalization, no AI, no 2 AM help
Implication: Represents "old model" Numi disrupts. Success validates demand for expert guidance.

B2B

SleepNest

Pricing: $197-397 (for consultants) | Market: Pediatric sleep consultants
Strengths: Clear B2B positioning, professional analytics
Implication: Validates B2B opportunity. Numi's Product 3 (Omnichannel Agent) goes further—handles communication, lead gen, after-care.

Tracker

Baby Connect & Others

Pricing: Free + $4.99 | Features: Manual tracking (feeding, sleep, diapers)
Weaknesses: Minimal AI, tracking only (no coaching), no sleep methodology
Implication: Shows demand for tracking but not competitive in coaching space.

Non-Tech Alternatives

Alternative	Cost	Limitations
Facebook parenting groups	Free	Unvetted advice, conflicting info, no personalization
Reddit communities	Free	Anonymous, inconsistent quality, no accountability
Baby sleep books	ILS 50-120	Static, not interactive, no 2 AM help
Private sleep consultants (Israel)	ILS 1,500-3,000	Expensive, limited availability, not scalable
Pediatrician advice	Free (with visit)	General, limited time, not sleep specialist

5.5 Israeli Market Analysis

Characteristic	Data	Implication for Numi
Fertility rate	2.9 children/family (2x OECD avg)	Higher demand per family; sibling logic critical
Annual births	~182,000	Structurally self-renewing market
WhatsApp usage	98% of population	Natural reach channel
Smartphone penetration	91%	Tech infrastructure ready
Active sleep consultants	Hundreds (no uniform regulation)	Unregulated market; parents seek alternatives
Sleep consulting cost	ILS 1,500-3,000 per package	Numi offers comparable annual support at same price
Health fund reimbursement	Up to 75% (up to 6mo postpartum)	Partnership opportunity with health funds
AI competitor in Hebrew	None	Numi will be first

Israeli Private Consultants vs Numi

Feature	Private Consultants	Numi
Price	ILS 1,500-3,000 per package	ILS 1,500-3,000/year
Availability	Business hours + phone	24/7
Response time	Hours to days	Seconds
Methodology	Varies between consultants	Dorit's method—uniform & proven
Ongoing guidance	Limited (2-4 calls)	Continuous throughout process
WhatsApp	Manual, not always available	Automatic, 24/7
Scalability	Limited to one consultant	Unlimited

5.6 SWOT Analysis (Israel Market)

Strengths

Agentic AI—autonomous with no human bottleneck
Native Hebrew (not translation, built from ground up)
WhatsApp integration (98% usage)
Dorit Kreizer's proven methodology
Competitive pricing (70-85% savings vs consultant)
3-product system (Lead gen + After-care + Omnichannel)
24/7 availability, response in seconds
Sibling logic (2.9 children/family)

Weaknesses

Early stage, no traction yet
No clinical research (planned Year 1)
Small team vs well-funded competitors
Brand awareness = zero
Dependence on single methodology
AI hallucination risk (mitigated with guardrails)
Higher price than global apps (justified by Hebrew/WhatsApp/proactivity)

Opportunities

Israeli market with no AI competitor
Partnership with health funds (kupot cholim)—75% reimbursement
Sleep consultant market as B2B channel (hundreds in Israel)
Global expansion after Israel success ($250M+ SAM)
Integration with Nanit (HQ in Tel Aviv, $70M+ funding)
Complementary services (nap training, regressions, sibling sleep)
Employee wellness programs (tech companies seeking parenting benefits)
Agentic AI going mainstream—consumer acceptance rising

Threats

Huckleberry adds Hebrew (low probability—small market for them)
Berry becomes agentic (requires architecture change)
Regulatory uncertainty (mitigated: advisory service, non-medical)
Privacy concerns (GDPR/HIPAA compliance, full transparency)
New Israeli competitors (first-mover advantage + Dorit partnership = moat)
Apple/Google native solution (unlikely in Hebrew)
Cultural resistance to AI (position as "Digital Dorit")
Economic downturn (baby sleep = critical pain; parents pay even in tough times)

5.7 Five Market Gaps Numi Fills

Gap	Current State	Numi's Solution
1. Omnichannel AI in Hebrew	No competitor operates AI agent on WhatsApp/Instagram/Facebook in Hebrew	Product 3 connects all channels from single AI brain
2. Proactive agentic capability	All competitors (including Berry) are reactive—wait for question	Numi operates autonomously: monitors, detects, acts, guides
3. B2B for sleep consultants	SleepNest offers analytics; no one offers autonomous agent	Products 1 & 2 serve consultants—lead gen & after-care management
4. Sibling logic	No competitor coordinates sleep between siblings	Coordination algorithm considers entire family schedule
5. Longitudinal memory	Competitors provide point-in-time recommendations	Numi tracks across full 0-24mo+ journey, across children in same family

5.8 Competitive Summary

The baby sleep consulting market is transforming. Huckleberry's Berry confirms the industry is moving toward conversational AI—but Berry is an answer to a question, not an agent that manages a process.

Numi enters with three core competitive advantages that are difficult to replicate:

Agentic architecture from ground up—not an upgraded chatbot, but an agent designed for autonomous operation
Native Hebrew + WhatsApp—a competitive moat in the Israeli market that global competitors won't invest in
Partnership with Dorit Kreizer—a proven methodology providing immediate credibility and clinical depth

In the Israeli market, Numi will have no direct competitor on launch day. In the global market, Numi competes on its agentic advantage—the difference between a tool that answers and an agent that leads.

6. Sales & Marketing Strategy

6.1 Overview

Numi's marketing strategy is built on one core principle: Leverage Dorit Kreizer's personal brand and existing community as the primary growth engine, supplemented by targeted paid marketing with a smart budget.

Parameter	Traditional Approach	Numi's Approach
Year 1 marketing budget	$100K-$150K	$60K ($5K/month)
Target market	US and English-speaking	Israel (Y1), UK (Y2), US (Y3)
Pricing model	Freemium (free + premium)	Annual subscription only, no free tier
Primary channel	Paid marketing	Organic through Dorit + supplementary paid
Revenue model	Direct B2C	50/50 partnership with Dorit
Price range	$12.99/month	$400-$800/year (ILS 1,500-2,960)

6.2 Year 1—Israeli Market

6.2.1 Dorit's Organic Channels (Cost: $0)

Numi's most significant marketing asset is Dorit Kreizer herself. Her organic channels generate high-quality exposure at zero marketing cost:

Social Media & Existing Clients

Active Instagram/Facebook with parent community
Existing client base and referrals
Every satisfied client is natural Numi advocate
Presented as continuation of personal consulting

Workshops & Events

Regular parenting workshops
Demonstrate Numi in action
Baby Fair, health fund workshops, community centers
Each event = opportunity to showcase product

Book Readers

"Shhh... At Night We Sleep" published by Kinneret Zmora
Sold at Steimatzky, Tzomet Sfarim, largest baby retail chain
Every book buyer is warm lead
Strategy: Add QR code in next edition linking to Numi

Israeli Parenting Communities

Facebook groups: "Israeli Moms," "Parents Talk," "New Moms"
WhatsApp neighborhood/daycare/birth prep groups
Forums: Tapuz Parents, BabyCenter Israel
Dorit (or recommendations about her) already present

Podcasts & Media

"The Parents Of," "Immaleh," other parenting/health podcasts
Each appearance = targeted audience exposure
Dorit talks about Numi as natural part of professional story

Professional Network

Lactation consultants, doulas, developmental therapists
Mutual referrals
Professional network increases exposure at no cost

Organic Channel	Monthly Leads	Conversion Rate	Paying Customers/Month
Dorit's social media	200-400	8-12%	16-48
Referrals from existing clients	50-100	15-25%	8-25
Workshops and events	30-60	10-20%	3-12
Parenting communities	100-200	5-10%	5-20
Book readers	50-150	8-15%	4-23
Podcasts and partnerships	50-100	5-10%	3-10
Total Organic	480-1,010	8-15% avg	39-138

Organic channels alone can generate 39-138 paying customers per month at zero marketing cost. This is the foundation of the entire strategy.

6.2.2 Paid Marketing ($5,000/month = $60K/year)

Channel	Monthly Budget	% of Budget	Objective
Facebook/Instagram ads	$2,500	50%	Targeted ads to Israeli parents, Reels/Stories campaigns
Israeli Facebook groups (paid)	$750	15%	Paid presence in large parenting groups
Google Ads (Hebrew keywords)	$1,250	25%	"Sleep consultant," "baby won't sleep," "baby sleep training"
Sponsored Instagram content	$500	10%	Reels/Stories demonstrating Numi in action
Total Monthly	$5,000	100%	—
Total Annual	$60,000	—	—

Facebook/Instagram Ads ($2,500/month)

Target: Israeli parents of 0-24mo babies, age 25-40
Formats: Stories ads (high CTR), sponsored Reels, carousel
Key messages: "It's 02:00 AM. Baby crying. You have Dorit in your pocket"
Expected CPM: $6-$10 (significantly lower than US $20+)
Monthly exposure: 250,000-400,000

Google Ads—Hebrew ($1,250/month)

Key advantage: Competition for Hebrew keywords significantly lower than English. CPC 50-70% lower than US.

Keyword (Hebrew)	Monthly Searches	Expected CPC
"Sleep consultant"	2,000-4,000	$0.50-$1.00
"Baby won't sleep"	3,000-6,000	$0.30-$0.70
"Baby sleep training"	1,000-2,500	$0.40-$0.80
"Sleep regression"	500-1,500	$0.30-$0.60

6.2.3 Pricing Strategy (No Freemium)

Package	Annual Price (USD)	Annual Price (ILS)	What's Included
Basic	$400/year	~ILS 1,500	AI chat based on Dorit's method, personalized sleep plan, sleep tracking, weekly reports
Pro	$750/year	~ILS 2,775	Everything in Basic + advanced 24/7 AI guidance, regression alerts, adaptive plans, WhatsApp support
Premium	$800/year	~ILS 2,960	Everything in Pro + one personal consultation call with Dorit, priority support, exclusive content

Pricing Justification—Comparison

Alternative	Cost	What You Get	Limitations
Private consulting (basic)	ILS 1,500-2,000	1-2 sessions + plan	Time-limited, no 24/7
Private consulting (extended)	ILS 2,500-3,000	3-4 sessions + weekly guidance	Still time-limited, not at 2 AM
Recorded sleep course	ILS 200-500	Static content	No interaction, no personalization
Numi Basic	ILS 1,500/year	24/7 AI for full year	Full Dorit's method access
Numi Pro	ILS 2,775/year	Advanced + WhatsApp	Everything + regressions

Core message: "For the price of one sleep consultation, you get a full year of 24/7 guidance from Dorit's method—available exactly when you need it, at 2 AM."

Why No Freemium Model

High value justifies payment—exhausted parents willing to pay for real solution
Predictable revenue—annual subscription upfront creates stable cash flow
Customer quality—paying customers more committed and successful
50/50 model—partnership with Dorit requires clear revenue from each customer
Small Israeli market—in 180,000 births/year market, paying customers preferable to free users

6.2.4 50/50 Partnership Model

Revenue Structure: Dorit refers clients → Client signs up for Numi → Revenue split 50/50

Example: Client pays $600/year (Pro) → $300 to Dorit, $300 to company

Incentive for Dorit: Passive income from every referred client—no additional work
Incentive for company: Dorit actively promotes Numi (serves her financial interest)

Metric	Conservative	Base	Optimistic
Paying customers (end Year 1)	200	400	700
Avg revenue/customer/year	$500	$600	$650
Annual gross revenue	$100K	$240K	$455K
Company share (50%)	$50K	$120K	$228K
Dorit's share (50%)	$50K	$120K	$228K

6.2.5 Customer Acquisition Funnel

Awareness: 300K-500K/month (Dorit + paid ads)

Website/Demo: 3,000-5,000 (1% of exposure)

Qualified Leads: 300-500 (10% of visits)

Purchase: 30-60/month (10% of leads)

Renewal: 60-70% (Year 2)

6.3 Year 2—UK Expansion

6.3.1 Funding the Expansion

UK expansion funded from: (1) Profits from Israeli market Year 1 (base: $120K company share), (2) Remaining capital from initial loan/investment.

6.3.2 UK Budget Allocation

Item	Annual Budget	Notes
Paid advertising (Facebook/Instagram/Google UK)	$80K-$120K	Largest share of UK budget
Additional developer (part-time/freelance)	$30K-$50K	English localization + cultural adaptation + UK regulatory compliance
Increased cloud costs	$15K-$25K	Multi-region deployment, high traffic
British market consultant	$20K-$30K	Expert in British parenting culture, NHS guidelines, message adaptation
Total	$145K-$225K	—

6.3.3 UK Market Pricing

Package	Annual Price (GBP)	Comparison
Basic	GBP 300/year	UK consultant package: GBP 200-500
Pro	GBP 450/year	—
Premium	GBP 600/year	—

Same core message: "For the price of one consultation, get a full year of 24/7 AI-powered guidance."

6.3.4 UK Marketing Strategy

Channel	Description	Budget %
Facebook/Instagram Ads (UK)	CPM ~$10.85 (47% lower than US $20.48). Targeted campaigns for British parents	50%
Google Ads (UK)	Keywords: "baby sleep consultant UK," "sleep training," "baby won't sleep"	20%
British parenting influencers	Collaborations with British bloggers/Instagrammers	15%
Mumsnet & British forums	Largest UK parenting forum—organic and paid presence	10%
NHS partnerships & NCT	NCT (National Childbirth Trust) groups, health visitors	5%

6.4 CAC & LTV Analysis

$30-$80

Blended CAC (Israel)

$600

Company LTV (after 50/50 split)

7.5-20x

LTV:CAC Ratio

1-3mo

CAC Payback Period

Israeli Market

Metric	Value	Explanation
Organic CAC (through Dorit)	$0-$50	Minimal marginal cost—existing network does the work
Paid CAC (ads)	$50-$150	Depends on channel and seasonality
Blended CAC	$30-$80	Weighted average (majority organic)
Gross LTV	$1,200	$600/year avg × 2 years retention
Company LTV (after 50/50 split)	$600	$300/year × 2 years
LTV:CAC ratio	7.5x - 20x	Very healthy (benchmark: 3x+)
CAC payback period	1-3 months	Annual upfront subscription ensures immediate payback

LTV:CAC Ratio by Channel

Channel	CAC	Company LTV	LTV:CAC	Assessment
Referrals from Dorit	$0-$20	$600	30x+	Excellent
Customer referrals (word of mouth)	$0-$10	$600	60x+	Excellent
Workshops and events	$30-$50	$600	12-20x	Very strong
Facebook/Instagram Ads	$80-$120	$600	5-7.5x	Good
Google Ads (Hebrew)	$60-$100	$600	6-10x	Good
Facebook groups	$40-$80	$600	7.5-15x	Very strong

6.5 Key Performance Indicators (KPIs)

Metric	Definition	Year 1 Target	Year 2 Target
Cost per Lead (CPL)	Avg cost to acquire qualified lead	$5-$15	$8-$20 (UK)
Cost per Acquisition (CPA)	Avg cost to acquire paying customer	$30-$80	$50-$120 (UK)
Monthly Active Users (MAU)	Customers who used Numi in past month	250-500	800-1,500
Conversion Rate (Lead to Paid)	% of leads who became paying customers	8-12%	10-15%
Retention Rate	% of customers who renew annual subscription	— (first year)	60-70%
Net Promoter Score (NPS)	Customer satisfaction & willingness to recommend	>50	>60
WhatsApp Engagement Rate	% of customers active in AI chat	>70%	>75%
Referrals per Customer	Avg referrals each customer generates	0.5-1.0	0.8-1.2
Monthly Recurring Revenue (MRR)	Monthly revenue from subscriptions (company share)	$5K-$10K	$25K-$50K

6.6 Sales & Marketing Strategy Summary

Element	Year 1 (Israel)	Year 2 (Israel + UK)
Marketing budget	$60K ($5K/month)	$145K-$225K (including UK)
Primary channel	Organic through Dorit (60-70% of customers)	Organic in Israel + paid in UK
Secondary channel	Paid—FB/IG/Google (30-40%)	Expanding channels in both markets
Pricing model	Annual subscription: ILS 1,500-2,960	Annual: GBP 300-600 (UK)
Revenue model	50/50 with Dorit	50/50 in Israel, TBD in UK
Blended CAC	$30-$80	$50-$120 (UK)
Company LTV	$600 (after split)	$600-$900
LTV:CAC ratio	7.5-20x	5-12x
Paying customers (year-end)	200-400	800-1,500
MRR (company share)	$5K-$10K	$25K-$50K

The advantage of this approach: Instead of burning hundreds of thousands on paid marketing in a competitive US market, we leverage an existing asset (Dorit) in a market we know well (Israel), prove Product-Market Fit with minimal spend, and then expand to the UK with a proven model and existing revenue. This is true Lean Startup—not just in theory.

7. Management Team

7.1 Team Philosophy

Numi is built by two founders only — a deliberate choice that creates competitive advantages at the Pre-Seed stage: zero founder salary cost, rapid decision-making, total commitment, and absolute focus.

Core Principle: Numi requires exactly two things — advanced AI technology and proven sleep methodology. Rotem provides the first. Dorit provides the second. Everything else can be hired or purchased — but these cannot be faked.

Advantage	Explanation
Founder salary: $0	Both founders draw zero salary in Year 1
Rapid decisions	Two people = zero bureaucracy, direct communication, immediate execution
Total commitment	Both invest personal capital (money or time) — not just "employees"
Focus	No organizational politics, no "managing managers" — just building

7.2 Founding Team

Rotem Levi — Co-Founder, CTO

Current Role	Senior Software Engineer, Bigabid (AdTech, public company)
Ownership	50%
Salary (Year 1)	$0 — earns living from Bigabid
Time Commitment	15-20 hours/week (evenings, weekends, vacation days)

Technical Expertise:

Full-Stack Development (React, Node.js, Python)
AI/ML and LLM Integration (Google ADK, Gemini)
Cloud Architecture (GCP, Infrastructure-as-Code)
Data Engineering & Real-Time Systems
WhatsApp/Social API Integration

Responsibilities: Building complete AI Agent platform, RAG pipeline, WhatsApp/social integrations, technical infrastructure, AI quality control

Why Rotem is Critical: Without Rotem, there is no platform. Dorit's methodology remains a book, not a technology product.

Dorit Kreizer — Co-Founder, Chief Content Officer

Current Role	Baby sleep consultant, author of "Shhhh... At Night We Sleep"
Ownership	50%
Salary (Year 1)	$0 — Revenue Share model
Investment	$100,000 owner's loan to company

Areas of Expertise:

Infant sleep methodology (certified consultant)
Clinical sleep consulting (hundreds of families)
Content creation & training data development
Client management & UX insight
Marketing & brand (existing follower base)
Business development & partnerships

Responsibilities: Proprietary methodology powering AI, training data creation, AI response validation, client relationships, organic marketing, B2B partnerships

Why Dorit is Critical: Without Dorit, Numi is a generic chatbot. Dorit's methodology is the irreplicable competitive moat.

7.3 Why This Team Works

Founder Salary Cost (Year 1)

50/50

Equity Split — Equal Partnership

$100K

Owner's Loan Investment

14 months

Runway with Zero Revenue

Complementary Skills Matrix

Domain	Rotem	Dorit
Technology & AI	LEADS	—
Sleep Methodology	—	LEADS
AI Agent Development	LEADS	Validates
Training Data Creation	Implements	CREATES
Client Relationships	—	LEADS
Marketing & Social	—	LEADS
Technical Infrastructure	LEADS	—
AI Response Quality	Implements	VALIDATES
B2B Partnerships	Supports	LEADS
Financial Management	SHARED	SHARED
Strategic Decisions	SHARED	SHARED

Technology without methodology = generic chatbot (no value)
Methodology without technology = consulting limited to one person (no scale)
Technology + methodology = AI platform with proprietary methodology (unique, scalable value)

7.4 Equity Structure & Governance

50/50 Ownership Rationale

Contribution	Rotem (50%)	Dorit (50%)
What they bring	AI technology, platform development	Proprietary methodology, brand, clients
Monetary value	Work valued at $96K-$144K/year (at market rate)	$100K loan + methodology + client base
Without them	No platform, no technology product	No methodology, no content, no clients
Outcome alone	Generic chatbot (no value)	Limited consulting business (no scale)

Corporate Governance

Decisions Requiring Joint Approval

Expenditure >$1,000
Any new contractual commitment
Hiring employees or contractors
Strategic changes (pivot)
Additional fundraising
B2B partnerships

Control Mechanisms

Monthly: Financial review (revenue, expenses, cash flow)
Quarterly: Strategic review (KPIs, roadmap, adjustments)
Ongoing: Transparent reporting — both founders see every transaction

Founders' Agreement

Clause	Detail
IP Ownership	All IP created belongs to the company
Non-Compete	No competing product during activity term
Data Ownership	All customer data, code, databases belong to company
Dispute Resolution	Mediation before legal proceedings
Exit Clauses	Clear mechanism if founder wishes to leave
Vesting	4 years with 1-year cliff — protects both parties

7.5 Future Hires & Advisory Board

Future Hiring Plan

Period	Role	Type	Prerequisite
Year 2	Developer (Part-Time)	Contractor	English localization for UK expansion
Year 2	UK Market Advisor	Contractor	Local knowledge — regulation, culture, marketing
Year 2+	Customer Support	Contractor/Part-Time	When user base grows beyond AI capacity
After Rotem's transition	Rotem full-time CTO	Employee	Transition triggers met (see Section 12)

Hiring Principles: No speculative hires; contractors before employees; growth-focused; protect runway.

Advisory Board (In Progress)

Pediatric Sleep Medicine

Clinical credibility, regulatory navigation, methodology validation

0.25-0.5% equity

AI/ML Advisor

Architecture review, model safety, scaling guidance

0.25-0.5% equity

SaaS Growth Advisor

GTM strategy, B2B sales, Unit Economics optimization

0.25-0.5% equity

7.6 Summary — Investor FAQs

Question	Answer
Why only two people?	That is all needed at this stage. Technology + methodology = product.
Where is the CMO?	Dorit — with existing brand, client base, and social media presence.
Where is the full-time CTO?	Rotem — 15-20 hours/week now, transitioning to full-time when numbers prove it.
Why no development team?	Rotem is building MVP himself. Additional team hired when revenue justifies it.
Why no salaries?	Both founders can afford it — 100% of money goes to product.
What if one leaves?	Founders' agreement with vesting, exit clauses, company IP ownership.
Why 50/50?	Without technology: no product. Without methodology: no value. Both equal.

8. Operations Plan

8.1 General Overview

Numi's operations are built on: lean, fast, and data-driven. Instead of building a full application, we launch an AI agent inside WhatsApp — enabling rapid launch (3 months to MVP), low operating costs (~$7,000/month), and zero app-download friction.

Guiding Principle: WhatsApp-First, Not App-First. Launch in 3 months on a platform Israeli parents already use daily. Zero friction. Zero downloads.

8.2 Technology Stack

Layer	Technology	Rationale
AI Framework	Google ADK (Agent Development Kit)	Purpose-built for multi-agent systems, native Gemini integration
AI Model	Gemini 2.0 Flash	Cost-efficient, fast, strong Hebrew support, Google ecosystem
Primary Channel	WhatsApp Business API	>95% penetration in Israel, parents use daily
Web Interface	Next.js (React)	Dashboards, data visualization, admin panel
Database	PostgreSQL with pgvector	Relational data + vector embeddings for RAG
RAG Pipeline	Custom-built on Google Cloud	Retrieval from Dorit's methodology knowledge base
Cloud Infrastructure	Google Cloud Platform (GCP)	Full integration with ADK and Gemini
Monitoring	Google Cloud Monitoring + Logging	Native integration, cost-effective
CI/CD	GitHub Actions	Automated deployment, automated testing

8.3 Why Google ADK + Gemini (Not OpenAI)

1. Native Multi-Agent Support

ADK designed for orchestrating multiple agents (sleep, nutrition, siblings, memory). Built-in routing, context management, error handling. OpenAI requires custom orchestration — adding 2-3 months development time.

2. Hebrew Language Support

Gemini demonstrates strong Hebrew performance, including cultural contexts and local expressions. Critical for speaking to Israeli mothers at 2 AM like an experienced consultant, not a translation engine.

3. Cost Efficiency

Model	Cost (500 clients)
Gemini 2.0 Flash	$150-300/mo
GPT-4o	$3,000-6,000/mo

Potential difference: 10-20x in AI cost. For a startup with $7,000 monthly burn, this is the difference between viability and runway threat.

4. Unified Ecosystem

ADK + Gemini + Google Cloud = seamless ecosystem. Data flows from WhatsApp through ADK to Gemini and back, with all components communicating natively. No translation layers between different cloud providers.

8.4 System Architecture

                    +-------------------+
                    |   WhatsApp API    |
                    | (Primary          |
                    |  Communication    |
                    |  Channel)         |
                    +---------+---------+
                              |
                    +---------+---------+
                    |   Google ADK      |
                    |   Agent Router    |
                    | (Primary Agent    |
                    |   Router)         |
                    +---------+---------+
                              |
          +----------+--------+--------+----------+
          |          |                 |           |
    +-----+------+  +-------+---+  +--+--------+  +------+-------+
    | Sleep Agent |  | Nutrition  |  | Siblings  |  | Memory Agent |
    |            |  | Agent      |  | Agent     |  | Long-Term    |
    |            |  |            |  |           |  | Memory       |
    | P0 - MVP   |  |            |  |           |  |              |
    +-----+------+  +-------+---+  +--+--------+  +------+-------+
          |          |                 |           |
          +----------+--------+--------+----------+
                              |
                    +---------+---------+
                    |   Gemini 2.0      |
                    |   Flash + RAG     |
                    +---------+---------+
                              |
              +---------------+---------------+
              |                               |
        +-----+--------+            +--------+--------+
        | PostgreSQL    |            | Dorit's         |
        | + pgvector    |            | Knowledge Base  |
        | (User Data)   |            | (Book, Content, |
        |               |            |  Methodology)   |
        +--------------+            +-----------------+

Agent Descriptions

Sleep Agent (P0, MVP)

Core agent: baby profile collection, personalized sleep plan, real-time guidance, progress monitoring, plan adjustment

Nutrition Agent (P1, Phase 2)

Feeding-sleep connection: tracking routines, night feeding recommendations, synchronizing feeding/sleep times, distinguishing hunger from habit

Siblings Agent (P1, Phase 3)

Multi-child families: parallel schedules, room-sharing, mutual disturbances, coordinating transitions

Memory Agent (P1, Phase 3)

Long-term context: tracking development, identifying regressions, adapting to developmental stages, cumulative knowledge

Architectural Principles

WhatsApp-First: Entire UX runs through WhatsApp. No app download, no new account, no new interface.
RAG-Based Knowledge: AI responses from Dorit's knowledge base, not model's "general knowledge." Ensures accuracy, prevents hallucinations.
Stateful Conversations: Full context maintained. AI remembers what was said a week ago, what the plan is, progress.
Graceful Escalation: When AI uncertain (low confidence, medical case, unusual situation) — auto-escalates to Dorit.
Cost-Aware Design: Every decision examined through cost lens. No room for waste at $7,000/month.

8.5 Development Roadmap

Phase 1: MVP

Months 1-3
First paying customer
Prove product works

Phase 2: Product-Market Fit

Months 3-6
25-50 customers
Validate PMF

Phase 3: Growth Features

Months 6-12
100-200 customers
Build retention

Phase 4: UK Expansion

Months 9-12
UK infrastructure
Localization

Phase 1: MVP (Months 1-3)

Feature	Description	Priority
WhatsApp Bot	AI conversation through WhatsApp Business API	P0
Sleep Agent	Consulting based on Dorit's methodology	P0
RAG Pipeline	Knowledge retrieval from Dorit's content (book, lectures)	P0
Intake Questionnaire	Baby profile + sleep history through natural conversation	P0
Payment Integration	Annual subscription via website, link from WhatsApp	P0
"Dorit's Clock"	Response delay timer — managing wait periods before room entry	P0

Milestones: Month 1: Infrastructure, WhatsApp API, RAG build | Month 2: End-to-end sleep agent, intake, Dorit testing | Month 3: 10+ beta users, bug fixes, limited launch

Phase 2: Product-Market Fit (Months 3-6)

Feature	Description	Priority
Nutrition Agent	Tracking feeding-sleep connection	P1
Web Dashboard	Sleep data visualization for parents	P1
Proactive Alerts	AI messages based on identified patterns	P0
Sleep Plan Generator	Personalized multi-week sleep plan	P0
Review Panel for Dorit	Dashboard for edge cases, escalated conversations	P1

Goal: 25-50 paying customers, AI accuracy >90%

Phase 3: Growth Features (Months 6-12)

Feature	Description	Priority
Siblings Agent	Multi-child sleep coordination	P1
Long-Term Memory	Cross-month developmental tracking	P1
Regression Prediction	Identifying sleep regressions before they occur	P1
Referral Mechanism	Built-in referral system for existing parents	P1
Analytics Dashboard	Business metrics: revenue, churn, usage	P1

Goal: 100-200 customers, all four agents active, referral mechanism

Phase 4: UK Expansion (Months 9-12)

Feature	Description	Priority
English Localization	Full platform translation + cultural adaptation	P1
Multi-Region Deployment	Google Cloud UK region	P1
UK Pricing	Payment in GBP	P1
UK Compliance	GDPR compliance, local data storage	P0

Goal: UK infrastructure ready for launch

8.6 Performance Targets & Monitoring

<3s

AI Response Time

99.5%

System Uptime

>90%

AI Accuracy (Validated by Dorit)

>99%

WhatsApp Delivery Reliability

>50

Customer Satisfaction (NPS)

<$1

AI Cost per Client/Month

Monitoring Tools

Tool	Function
Google Cloud Monitoring	System performance, uptime, latency
Google Cloud Logging	Detailed logs of every conversation and action
PostgreSQL Analytics	Usage metrics, customer behavior, retention
Custom Dashboard (Next.js)	Business metrics: revenue, clients, CAC, LTV

8.7 Security & Privacy

Domain	Measure	Status
Encryption in Transit	TLS 1.3 for all communications	Built into GCP and WhatsApp
Encryption at Rest	AES-256 for all data in PostgreSQL	Built into Google Cloud SQL
Israeli Privacy Law	Compliance with Privacy Protection Law, 5741-1981	In planning, pre-launch
GDPR	Full compliance (for UK expansion)	In planning, pre-launch
No Data Sharing	No sharing of personal info with third parties	Fixed policy
Right to Deletion	Parents can request data deletion anytime	In development
AI Safety Guardrails	Medical content filtering, auto-escalation in edge cases	In development

Medical Case Handling: Numi is not a medical product. In cases with signs of medical issues, psychological distress, or complex medical conditions, AI identifies and escalates immediately to Dorit or refers to doctor/emergency hotline.

8.8 Key Milestones (24 Months)

Month 1-3

Milestone: MVP on WhatsApp with Sleep Agent

Success Metric: 10+ beta users, AI accuracy >85%

Month 3-6

Milestone: First paying customers

Success Metric: 25-50 paying customers, satisfaction >80%

Month 6-9

Milestone: Nutrition and Siblings agents

Success Metric: 100-200 customers, AI accuracy >90%

Month 9-12

Milestone: UK expansion preparation

Success Metric: 200-500 customers, English localization complete

Month 12 ★ Break-Even: 280 Customers

Milestone: Operational break-even — revenue covers costs

Success Metric: 280+ active customers, $7,000+ monthly revenue

Month 12-18

Milestone: UK launch + growth

Success Metric: 400-1,000 customers (Israel + UK)

Month 18-24 ★ Scale + Loan Repayment

Milestone: Scale, loan fully repaid

Success Metric: 800-2,000 customers, loan fully repaid

8.9 Technology Risk Management

Risk	Probability	Impact	Mitigation
Google changes ADK/Gemini pricing	Medium	High	Modular architecture enables switch to another model (Claude, Llama) within weeks
WhatsApp Business API restrictions	Low	High	Building infrastructure for alternative channels (Telegram, SMS, Web Chat) early
Data breach	Low	Very High	Encryption at all layers, principle of least privilege, monthly access audit
AI gives incorrect advice	Medium	High	RAG reduces hallucinations, clear disclaimers, escalation mechanism, sample auditing
Slow response times	Low-Medium	Medium	Smart caching, prompt optimization, CDN

9. Financial Plan & Projections

Financial Executive Summary

Numi is built as a lean startup with a conservative investment structure: $100K owner's loan from Dorit, with 50/50 ownership split. This structure allows full control, clear path to profitability, and relatively low risk.

$100K

Total Investment Required

~$7K

Monthly Burn Rate

~14 mo

Zero-Revenue Runway

280

Break-Even Customers

9.1 Investment Structure

Owner's Loan: $100K

Detail	Description
Funding Type	Owner's Loan from Dorit to company
Amount	$100,000 (~370,000 NIS)
Repayment Terms	From company profits, first priority before dividend distribution
Ownership Split	50% Rotem / 50% Dorit
Database & IP	Belong to company; Dorit's methodology licensed to company
External VC	None — zero dilution, full control

Why Owner's Loan vs. Venture Capital?

Criterion	Owner's Loan ($100K)	Venture Capital ($500K)
Ownership dilution	0% — 100% ownership retained	16.7% to external investors
Control	Full — both partners only	Board of directors, investor veto rights
Growth pressure	None — healthy pace	Aggressive — demand 10x in 3-5 years
Personal risk	Limited — $100K only	High — expectations for $5M+ ARR
Time to revenue	3-4 months (existing audience)	6-9 months (building from scratch)
Investment return	First priority from profits	Only at exit (sale/IPO)

9.2 Monthly Burn Rate (Year 1)

Category	Monthly (NIS)	Monthly ($)	Annual ($)	Notes
Google Cloud + Gemini API	1,850-3,700	$500-$1,000	$6K-$12K	Scales with users
WhatsApp Business API	740-1,850	$200-$500	$2.4K-$6K	Per-message pricing
Marketing & Advertising	18,500	$5,000	$60K	Facebook, Instagram, Google, content
Domain, Hosting, Tools	740	$200	$2.4K	Vercel, GitHub, dev tools
Legal & Accounting	1,850	$500	$6K	CPA, attorney, company formation
Rotem's Salary	0	$0	$0	Employed at Bigabid
Dorit's Salary	0	$0	$0	Revenue Share model
Total	~23,700-26,600	~$6,400-$7,200	~$77K-$86K

Use of Funds Breakdown

The Critical Number: 14-Month Runway

$100,000 / $7,000 per month = ~14.3 months

Even if not a single dollar comes in, the company has nearly 1.5 years of survival runway. But revenue expected from Month 3-4 thanks to Dorit's existing client base.

Scenario	Capital Burned	Capital Remaining
Conservative (revenue from M6)	~$42K	~$58K
Base (revenue from M4)	~$28K	~$72K
Optimistic (revenue from M3)	~$21K	~$79K

9.3 Revenue Model

Pricing & Payment Structure

Annual upfront payment (not monthly subscription). Payment split 50/50 between consultant (Dorit) and platform (Numi).

Package	Annual Price ($)	Annual (NIS)	Company Share ($)	Company Share (NIS)	What's Included
Basic	$400	~1,500	$200	~740	Single parent, one child, sleep agent only
Pro	$750	~2,775	$375	~1,388	Two parents, professional agent (intern/trainee), sleep + nutrition
Premium	$800	~2,960	$400	~1,480	Up to 3 children (siblings), full family + team, all agents + memory
Average	$600	~2,220	$300	~1,110	= $25/month per client to company

Why Annual Upfront Payment?

Advantages

Positive cash flow — money upfront
High commitment — client paid $600, committed to success
Low churn — no monthly exit points
Financial planning — revenue known in advance
High perceived value — $600/year = $50/mo, 70% cheaper than private consultant

Per-Client Revenue Breakdown

Client pays $600/year (avg)
         |
    +----+----+
    |         |
   50%       50%
    |         |
 $300       $300
 To         To the
 Consultant Company
            |
      $25/month/client
      = Net revenue

9.4 Break-Even Analysis

$84K

Annual Expenses

$300

Revenue per Client/Year (Company Share)

280

Clients Needed for Break-Even

2.6%

Of Potential Market (10,800 Israeli parents)

Break-Even: Revenue vs Expenses

When Does Break-Even Occur?

Scenario	Month Reaching 280 Clients	Capital Burned	Capital Remaining
Conservative	Month 12	~$60K	~$40K
Base	Month 9	~$45K	~$55K
Optimistic	Month 6	~$30K	~$70K

9.5 Three-Year Projections (Three Scenarios)

Revenue Growth: Three Scenarios

Conservative Scenario

Slow growth, low conversion, limited marketing, strong competition

Metric	Year 1 (Israel)	Year 2 (Israel + UK)	Year 3 (+ US)
Paying Clients (end of year)	100	400	1,200
Avg Price (company share)	$300/year	$300/year	$350/year
Annual Revenue	$30K	$120K	$420K
Annual Expenses	$84K	$180K	$350K
EBITDA	-$54K	-$60K	+$70K

Base Case Scenario

Reasonable growth, Dorit's existing client base, targeted marketing, UK entry Year 2

Metric	Year 1 (Israel)	Year 2 (Israel + UK)	Year 3 (+ US)
Paying Clients (end of year)	300	1,200	3,500
Avg Price (company share)	$300/year	$325/year	$350/year
Annual Revenue	$90K	$390K	$1.23M
Annual Expenses	$84K	$220K	$500K
EBITDA	+$6K	+$170K	+$725K

Optimistic Scenario

Dorit as strong brand, virality in parenting communities, Premium pricing works, UK succeeds quickly

Metric	Year 1 (Israel)	Year 2 (Israel + UK)	Year 3 (+ US)
Paying Clients (end of year)	500	2,000	6,000
Avg Price (company share)	$325/year	$350/year	$375/year
Annual Revenue	$163K	$700K	$2.25M
Annual Expenses	$86K	$300K	$700K
EBITDA	+$77K	+$400K	+$1.55M

Scenario Comparison

Metric	Conservative	Base	Optimistic
Year 1 Clients	100	300	500
Year 3 Clients	1,200	3,500	6,000
Year 1 Revenue	$30K	$90K	$163K
Year 3 Revenue	$420K	$1.23M	$2.25M
Year 3 EBITDA	+$70K	+$725K	+$1.55M
Loan Repayment	Year 3	Year 2	Year 1
Profitability	Year 3	End of Year 1	Year 1
ROI on $100K	x0.7 (Y1) → x4.2 (Y3)	x0.9 (Y1) → x12.2 (Y3)	x1.6 (Y1) → x22.5 (Y3)

9.6 Profitability & ROI

Three-Year P&L Overview

Return on Investment for Dorit

Scenario	Loan Repaid?	When?	Value of 50% Ownership (Year 3)	Total ROI
Conservative	Yes	Year 3	~$500K-$1M	x5-x10
Base	Yes	Year 2	~$1.5M-$3M	x15-x30
Optimistic	Yes	Year 1	~$3M-$6M	x30-x60

Asymmetric Bet

Downside: Maximum loss of $100K (~370K NIS) — and only if everything fails completely
Upside: 50% ownership in company generating $1.2M+ revenue in Year 3 (base case). Ownership value: $1.5M-$3M+
Risk/Reward Ratio: $100K → $1.5M-$3M = x15-x30

9.7 Sensitivity Analysis

Variable	Change	Impact on Break-Even	Impact on Year 1
Avg price drops to $400 (from $600)	Company share = $200	420 clients (vs 280)	Larger loss, but within runway
Marketing expenses x1.5	$90K annual (vs $60K)	380 clients	Additional loss of ~$30K
Growth 50% slower	150 clients (vs 300) in Y1	Break-even at Month 18	Loss of ~$39K, $61K remaining
Gemini API cost doubles	$18K annual (vs $9K)	300 clients	Small impact — additional $9K
Disaster scenario (all above)	All variables negative	~500 clients	Loss of ~$70K; $30K remaining; 4+ months runway

Even in worst-case scenario, company retains $30K and several months to correct course or wind down orderly. Maximum loss is $100K — likely much lower.

9.8 Key Assumptions & Sources

Financial Assumptions

Assumption	Value	Basis
Average price per client	$600/year	Average between Basic ($400) and Premium ($800)
Revenue split	50/50 (consultant/platform)	Standard in SaaS-for-experts market
USD/NIS exchange rate	3.70	Average rate 2025-2026
Gemini API cost per conversation	~$0.01-$0.05	Google Gemini 2.0 Flash pricing
WhatsApp cost per message	~$0.005-$0.05	Meta Cloud API pricing (varies by type)
CAC Israel	~$50-$100	Targeted digital marketing + referrals
CAC UK	~$100-$200	New market, awareness from scratch
Annual customer retention	~70%	Annual subscription renewals

Market Data Sources

Data Point	Source
Baby sleep apps market: $420M (2024)	Grand View Research, Allied Market Research
CAGR 12-13%	Grand View Research
60% of parents struggle with sleep	American Academy of Pediatrics
180,000 annual births in Israel	Israel Central Bureau of Statistics
~700K annual births in UK	ONS (Office for National Statistics)

9.9 Summary — Why This Investment Makes Sense

1. Capital Protection

Monthly burn only $7K = 14 months runway with $0 revenue
Revenue begins Month 3-4 thanks to existing audience
Actual risk: far less than $100K

2. Asymmetric Bet

Downside: limited to $100K (owner's loan, not "thrown away")
Upside: 50% ownership in multi-million dollar potential company
Risk/reward ratio: 1:15 to 1:60

3. Revenue from Day 1

Dorit already active consultant with client audience
Product accelerates and scales existing business
Not startup from zero — upgrade of existing business

4. Conservative Numbers

Break-even: only 280 clients
Market: 10,800+ Israeli parents willing to pay
280/10,800 = less than 3% penetration

5. Fair Structure

50/50 — equal partnership
Loan repaid with first priority
Data belongs to company
No external dilution

Question	Answer
How much money needed?	$100K (~370,000 NIS)
Worst case scenario?	Lose part of $100K, but not all
Best case scenario?	50% ownership in company worth $3M-$10M+
When is loan repaid?	12-24 months (depends on scenario)
When profitability reached?	End of Year 1 (base), Year 3 (conservative)
Who is in control?	Rotem and Dorit — 50/50, no external investors

10. Risk Analysis and Mitigation Plan

Guiding Principle: Asymmetric Bet
This is not an "all or nothing" bet. This is an asymmetric bet — a situation where the downside is limited and predefined ($100K), while the upside is open and expansive (50% ownership in an AI platform in a $1.25B market).

10.1 Risk 1: Technology Risk — AI Quality

⚙️

Risk Assessment

Probability	Low
Impact	High
Urgency	Months 1-6

🛡️

Mitigation

POC is working — proven on Google ADK + Gemini
RAG from Dorit's methodology — not generic AI
Dorit validates — ongoing review in Year 1
Safety guardrails — medical escalation, filtering
Mature tech — Google ADK proven infrastructure

10.2 Risk 2: Market Risk — Not Enough Customers

180,000

Annual births in Israel

280

Customers needed to break-even

0.16%

Market penetration required

📊

Why Impact is Medium (Not High)

Low operating costs ($7K/month) — long runway even with slow growth
Revenue begins early thanks to Dorit's existing client base
Israeli market is starting point — UK and US expansion planned
No AI competitor in Hebrew — Blue Ocean market
$60K/year marketing budget supplements organic growth

10.3 Risk 3: Competitive Risk — Huckleberry Enters Hebrew Market

🏁

Assessment

Probability	Low (1-2 year window)
Impact	Medium
Urgency	Year 2+

🚀

Numi's Advantages

Israeli market too small for global competitors
Hebrew is more than translation — RTL, cultural nuances
WhatsApp integration — Israel is WhatsApp-first
Dorit's recognized methodology vs. generic approach
First-mover advantage in small market is strong
Agentic AI vs. Reactive AI — architectural difference

10.4 Risk 4: Founder Risk — Rotem's Limited Time While Employed

⏱️

Mitigation

MVP can be built evenings/weekends — 15-20 hrs/week sufficient
Google ADK reduces complexity — weeks instead of months
Clear transition triggers — Rotem goes full-time at 1,000 customers (Section 12)
AI operates 24/7 — no human bottleneck after launch
Dorit covers all non-technical work — clean division of labor

10.5 Risk 5: Regulatory Risk — AI Health Consultation Regulation

⚖️

Assessment

Probability	Low (1-2 year window)
Impact	Low-Medium
Urgency	Year 2+ (mainly US)

✅

Protection

Numi provides parenting guidance — not medical advice
Clear disclaimers in every conversation
Automatic escalation for medical questions
Dorit's professional credentials add legitimacy
Israel's regulatory environment is permissive

10.6 Capital Protection — Five Scenarios

Scenario	Monthly Burn	Revenue Start	Loan Repayment	Capital at Risk	Remaining Value
Best Case	$7K/mo	Month 3	Months 12-14	$21K (~78K ILS)	$79K + revenue
Base Case	$7K/mo	Months 4-6	Month 25	$42K (~155K ILS)	$58K + revenue
Conservative	$7K/mo	Months 6-9	Months 30-33	$63K (~233K ILS)	$37K + revenue
Bad	$7K/mo	Month 12+	Month 30+	$84K (~311K ILS)	IP preserved
Catastrophic	$7K/mo	Never	Never	$100K (~370K ILS)	IP worth $55K-$120K

10.7 Why Even Worst-Case is Manageable

💼

IP Assets Preserved

Asset	Estimated Value
Numi Platform (code, architecture)	$30K-$60K
Codified Methodology (AI/content)	$10K-$20K
Intellectual Property Rights	$10K-$30K
Data Repository & Market Insights	$5K-$10K
Total	$55K-$120K

🛡️

Personal Protection

Maximum loss is known: $100K (~370K ILS) — no surprises
Dorit's consulting unaffected — continues as before
Rotem retains Bigabid position — no livelihood risk
No future obligations — no guarantees, no hidden debts
Owner's loan — internal debt, Dorit owns 50%

10.8 Why the Upside Justifies the Risk

$1.25B

Global market by 2033

50%

Ownership stake

10-20x

AI company revenue multiples

Active Customers	Annual Revenue	Company Share (50%)	Dorit's Ownership Share	Note
280	$168,000	$84,000	$42,000	Break-even
1,000	$600,000	$300,000	$150,000	Loan repaid + profit
3,500	$2,100,000	$1,050,000	$525,000	Significant profit
5,000	$3,000,000	$1,500,000	$750,000	Large profitable business

Company Valuation Potential (Year 3): At $1.2M ARR × 10-20x revenue multiple = $12M-$24M valuation. Dorit's 50% = $6M-$12M.

10.9 Risk-Adjusted Return Summary

Scenario	Probability	Outcome	Expected Value
Catastrophic (total loss)	5%	-$100K	-$5K
Bad (partial loss)	10%	-$50K	-$5K
Conservative (repayment + small profit)	25%	+$150K	+$37.5K
Base case (significant profit)	40%	+$500K-$1M	+$300K
Optimistic (large profit)	20%	+$2M-$5M	+$700K
Weighted Expected Value			+$1,027,500

Interpretation: For every dollar invested, the expected value (probability-weighted) is approximately $10 in return. This is an excellent risk/return ratio.

10.10 Summary Risk Matrix

Risk	Probability	Impact	Score	Primary Mitigation	Status
Technology (AI underperforms)	Low	High	Medium	POC working; RAG; Dorit validates	POC Active
Market (insufficient customers)	Medium	Medium	Medium	280 to break-even; existing audience; Blue Ocean	14mo runway
Competitive (competitor entry)	Low	Medium	Low-Med	First-mover; WhatsApp; methodology	Advantage
Founder (limited time)	Medium	Medium	Medium	Google ADK; triggers in Section 12	POC Built
Regulatory (AI regulation)	Low	Low-Med	Low	Parenting guidance; auto-escalation	In Progress

Bottom Line: Downside is limited to $100K (known, manageable, doesn't affect livelihoods). Upside is 50% ownership in an AI platform in a $1.25B+ market with 5x-50x ROI potential. This is a classic asymmetric bet — limited downside, open upside.

11. Loan Repayment Plan

Executive Summary

$100K

Loan Amount (~370K ILS)

$7K/mo

Monthly Operating Costs

14 months

Runway with zero revenue

280

Customers for break-even

📋

Loan Terms

Loan Type: Owner's Loan
Lender: Dorit Kreizer
Borrower: Agentic AI Ltd.
Interest Rate: 0%
Priority: Full priority over profit distribution (dividends) to shareholders
Repayment: From company profits, base case month 30-32, conservative month 36+

Revenue Structure

Component	Value
Annual price per customer (range)	$400-$800/year
Average price per customer	~$600/year (~2,220 ILS)
Split	50/50 (consultant/company)
Net revenue to company per customer	~$300/year (~1,110 ILS)
Monthly revenue to company per customer	~$25/month (~92.5 ILS)

Monthly break-even: $7,000 ÷ $25 = 280 active customers

Cash Flow Distribution Priority

Priority 1: Ongoing operating costs (~$7,000/month)

          |

          v

Priority 2: Owner's loan repayment ($100,000)

          |

          v

Priority 3: UK market expansion fund

          |

          v

Priority 4: Profit distribution to shareholders (dividends)

Implication: Every shekel of profit beyond operating costs goes to loan repayment first. Shareholder dividends begin only after full loan repayment.

Three Repayment Scenarios

Scenario A: Conservative

Metric	Value
Customers at Month 6	10
Customers at Month 12	50
Customers at Month 18	150
Customers at Month 24	280 (break-even)
Total cumulative deficit (months 1-24)	~$112,000
Required action	Reduce costs to ~$5,500/mo from month 15
Full repayment (projection)	Month 36+

Scenario B: Base Case

Metric	Value
Customers at Month 6	25
Customers at Month 12	100
Customers at Month 18	280 (break-even)
Customers at Month 24	490
Capital consumed for operations (months 1-18)	$81,600 (~302K ILS)
Cash reserve remaining	$18,400 (~68K ILS)
Cumulative repayment from surplus (months 19-24)	$17,375 (~64K ILS)
Loan balance at month 24	$82,625 (~306K ILS)
Full repayment (projection)	Month 30-32

Scenario C: Optimistic

Metric	Value
Customers at Month 6	50
Customers at Month 12	200
Customers at Month 14	280 (break-even)
Customers at Month 24	900
Capital consumed for operations (months 1-14)	$60,800 (~225K ILS)
Cash reserve remaining	$39,200 (~145K ILS)
Cumulative surplus months 15-24	$76,500 (~283K ILS)
Full repayment (projection)	Month 23

Scenario Comparison

Metric	Conservative	Base Case	Optimistic
Operational break-even	Month 24	Month 18	Month 14
Loan repayment begins	Month 25+	Month 19	Month 15
Full repayment (projection)	Month 36+	Month 30-32	Month 23
Maximum capital at risk	$100,000	$100,000	$100,000
Residual risk	Low	Low	Very low

Lender Protection Mechanisms

🔒

Protection 1: Asset Lien

Loan secured by lien on all company assets: Numi platform (code, architecture), data, IP, AI models, methodology rights. IP assets retain $55K-$120K value even in failure.

📊

Protection 2: Transparent Reporting

Monthly financial reports to all partners: P&L, cash flow, customer metrics, loan balance, repayment forecast, deviations with corrective steps.

✅

Protection 3: Expense Governance

Expenses >$1,000 require joint approval. All new contracts, salary changes, and marketing overruns need partner consent.

🏆

Protection 4: Liquidation Priority

In liquidation: 1) External debts, 2) Owner's loan ($100K), 3) Shareholder distribution. Loan is repaid before any dividends.

🗄️

Protection 5: Company Data Ownership

All code, data, IP owned by company (not founder personally). Assets secured within company, independent of individuals.

💵

Protection 6: Low Burn Rate

$7K/month burn rate is 4-14x lower than typical startups. Slow capital consumption = long runway to correct, pivot, or wind down.

👥

Protection 7: Existing Client Base

Dorit has active clients, book generates traffic, thousands of social followers, established reputation. Zero-revenue risk is extremely low.

Summary and Recommendation

Why This Loan is Safe: Relatively low amount ($100K), low burn rate ($7K/mo vs. $30K-$100K typical), full priority over dividends, asset lien on IP/code/data, transparent monthly reporting + joint expense approval, existing market (Dorit's clients/brand), corporate governance (expenses >$1K need approval), residual value ($55K-$120K even in failure).

Repayment Timeline:

Optimistic scenario: Full repayment Month 23
Base case scenario: Full repayment Month 30-32
Conservative scenario: Full repayment Month 36

Bottom line: This is a low-risk loan with multiple protection mechanisms, clear repayment path, and real assets securing the funds.

12. Founder Transition Plan

Current Status

Detail	Data
Current Employer	Bigabid (AdTech, publicly traded)
Role	Senior Software Engineer
Gross Salary	52,323 ILS per month
Net Salary (estimate)	~35,000 ILS per month
Salary Cost to Agentic AI	0 ILS — Rotem draws no salary from company
Time Dedicated to Numi	Evenings, weekends, vacation days

Implication: In Year 1, Rotem's salary cost is zero. The $100K loan goes entirely to product, marketing, and growth. Rotem supports his family from his current salary, allowing the company to run lean.

Transition Trigger: When Does Rotem Move to Full-Time?

Guiding Principle: Rotem will not leave his job based on hope. The transition occurs only when Numi can replace Rotem's income and sustain company operations. This conservative approach protects the company, family, and investment.

📐

The Formula

Transition Trigger: Company net monthly profit ≥ 2x Rotem's net salary

Rotem's net salary	~35,000 ILS per month
Safety multiplier	2x
Required net profit	~70,000 ILS/month = ~$18,900/month
Monthly operating expenses	~$7,000/month
Required monthly revenue	~$25,900/month

🎯

Why a 2x Multiplier?

Pay Rotem competitive salary (~52K ILS/month)
Continue covering operating expenses
Maintain growth pace (marketing + development)
Service existing obligations (loan repayments)
Provide safety cushion (no cash flow pressure)

Required Customer Count by Price Level

Price Level	Annual Price	Company Share (50%)	Monthly Rev/Customer	Customers for Transition
Basic ($400/year)	$400	$200	$16.67	~1,554
Pro ($750/year)	$750	$375	$31.25	~829
Premium ($800/year)	$800	$400	$33.33	~777
Weighted Avg ($600/year)	$600	$300	$25.00	~1,036

Bottom line: Once Numi has approximately 1,000 active paying customers at weighted average $600/year, Rotem can execute the transition.

Timeline Scenarios

Conservative Scenario: Transition Months 24-30

Growth rate: Slow — organic marketing primarily
Reach 1,000 customers: Month 24
Trend validation: Months 24-27 (3+ stable months)
Resignation submission: Month 27
Actual transition: Month 29-30
Risk: Very low — clear data before decision

Base Case Scenario: Transition Months 18-24

Growth rate: Moderate — paid marketing + Dorit's network
Reach 1,000 customers: Month 18-20
Trend validation: Months 18-21
Resignation submission: Month 21
Actual transition: Month 23-24
Risk: Low — aligned with break-even (month 18)

Optimistic Scenario: Transition Months 12-18

Growth rate: Fast — viral success, aggressive marketing
Reach 1,000 customers: Month 12-14
Trend validation: Months 12-15
Resignation submission: Month 15
Actual transition: Month 17-18
Risk: Low-Medium — data supports but shorter history

Risk Management — Five Layers of Protection

1️⃣

Proven Numbers Required

Transition contingent on at least 3 consecutive months of revenue above threshold. One good month isn't enough — stable, sustained trend required.

2️⃣

2x Safety Multiplier

Requirement for net profit double Rotem's salary provides wide cushion. Even if revenue drops 30% post-transition, company can still pay salary + expenses.

3️⃣

If Growth Stalls — Rotem Stays

No financial risk to company or family. If Numi doesn't hit threshold, Rotem continues at Bigabid, contributing in free time. Company operates without founder salary costs.

4️⃣

Dorit's Independence

Dorit continues independent sleep consulting. Her personal income doesn't depend on Numi's success. Both founders maintain financial stability even if delayed.

5️⃣

Zero Salary Obligations Until Transition

Until trigger is met, Agentic AI salary expenses are zero — both founders contribute without salary. This extends the loan runway significantly.

Work Arrangement Before Transition

👨‍💻

Rotem Levi (CTO / Co-Founder)

Evenings (2-3 hrs/day): Product development, AI architecture, team management
Weekends (6-8 hrs): Development sprints, code reviews, technical planning
Vacation days: Strategic tasks, investor meetings, product demos
Total: ~15-20 hours per week

👩‍⚕️

Dorit Kreizer (Sleep Expert / Co-Founder)

Consulting hours: Client sessions (personal income + data for AI)
Content creation: Writing, recordings, marketing materials
AI training: Reviewing/validating AI responses, training data
Marketing: Social media, professional network promotion
Total: ~15-25 hours per week

Combined contribution: ~30-45 hours/week at no salary cost. Market equivalent: $8K-$12K/month in personnel costs saved.

Pre-Transition Indicators

Indicator	Minimum Threshold	Ideal Threshold
Active paying customers	800+	1,200+
MRR	$20,000+	$30,000+
Monthly operating profit	$15,000+	$20,000+
Revenue trend (3 months)	Rising or stable	Consistently rising
Monthly churn rate	<7%	<5%
LTV:CAC	>5x	>10x
Number of B2B partnerships	1+	3+

Iron rule: All indicators must be above minimum threshold, and at least 4 out of 7 above ideal threshold, before Rotem submits resignation.

Why This is Smart: Summary for Investor

Rotem is not leaving his job based on hope. He is building measurable triggers that prove the business works before he makes the leap.

💰

1. Zero Financial Risk Year 1

CTO salary cost = $0. All loan funds go to product and growth. Extends runway, increases ROI.

🤝

2. Skin in the Game

Rotem dedicates evenings/weekends because he believes in the product. Highest level of founder commitment.

📊

3. Data-Driven Transition

Triggers based on numbers: revenue, customers, profitability. Not "I feel it's time" — "the numbers prove it's time."

🛡️

4. Full Protection for All

Company protected (no unsustainable salary), family protected (no financial risk), Dorit protected (loan funds not to salary too early).

🚀

5. Ready for Acceleration

Once full-time, development pace doubles. From ~15-20 hrs/week to 40-50 hrs. Accelerates roadmap, shortens time to V2.0/V3.0, strengthens competitive position.

13. Appendices

Appendix A: Glossary

Term	Explanation
Agentic AI	AI that operates autonomously — not just responding to questions, but initiating actions, monitoring, deciding, and executing tasks independently. Numi's fundamental differentiation vs. competitors.
RAG	Retrieval-Augmented Generation — AI technique where model searches defined knowledge base (Dorit's methodology) for relevant info, then builds personalized answer. Reduces hallucinations, ensures accuracy.
Google ADK	Agent Development Kit — Google's platform for building multi-agent systems. Provides infrastructure for agent coordination, memory, tool execution, query routing. Numi's tech foundation.
Gemini 2.0 Flash	Google AI model — fast, cost-effective, accurate. Selected for strong Hebrew support, <3 sec response time, cost 10-20x lower than GPT-4o.
Owner's Loan	Loan from shareholder (Dorit) to company (Agentic AI). No dilution. Repaid from profits with first priority before dividends. Amount: $100K (~370K ILS).
Revenue Share	Business model splitting customer revenue 50/50 between consultant (Dorit) and company (Agentic AI). Example: $600/year = $300 Dorit, $300 company.
Vesting	Founder shares vest gradually over 4 years. Protects company if founder leaves early. 1-year cliff — leaving before 1 year forfeits all shares.
EBITDA	Earnings Before Interest, Taxes, Depreciation, Amortization. Standard measure of operational profitability.
CAC	Customer Acquisition Cost — average cost to acquire one paying customer. Numi's weighted CAC (Israel): $30-$80 (low, thanks to organic channels).
LTV	Lifetime Value — total revenue customer generates over usage period. Numi gross LTV: $1,200 ($600/yr × 2 yrs). Net LTV to company: $600 (50/50 split).
MRR	Monthly Recurring Revenue — normalized monthly income from active subscriptions. MRR = customers × monthly revenue/customer. Numi: $25/customer/mo (company share).
ARR	Annual Recurring Revenue. ARR = MRR × 12. Used for SaaS valuations.
TAM/SAM/SOM	Total/Serviceable/Obtainable Addressable Market. TAM (3 markets): $1.71B. SAM (Year 3): ~$353M. SOM (Year 3 base): $1.22M.
Break-even	Point where revenue = expenses. Numi: 280 active customers (monthly revenue $7K = covering operational expenses).
Burn Rate	Rate of capital consumption per month. Numi: ~$7K/month (~25,900 ILS). 4-14x lower than typical startups.
Runway	Survival time with zero revenue. Numi: ~14 months ($100K ÷ $7K/month).
Churn Rate	% of customers canceling subscription in given period. Numi expected: <5%/month (industry avg: 8-12%), thanks to long-term memory moat.
MVP	Minimum Viable Product — first version with sufficient features to serve paying customers and validate assumptions. Numi MVP: sleep agent on WhatsApp with RAG.
POC	Proof of Concept — initial model proving tech works. Numi POC: operational on Google ADK + Gemini 2.0 Flash, demonstrates personalized conversations.
Blue Ocean	Market with no direct competition (vs. Red Ocean: fierce competition). Israeli AI-based sleep consultation is Blue Ocean — no AI competitor in Hebrew.

Appendix B: Data Sources

Source	Data Used	Section
Grand View Research	Baby sleep apps market: $420M (2024) → $1.25B (2033), CAGR 12-13%	Sections 4, 9
Allied Market Research	Broader baby sleep products: $6.5B (2023) → $10.2B (2032), CAGR 5.4%	Section 4
CBS Israel	181,609 births/year (2024); fertility 2.9; Q4 2024 baby boom +10%	Sections 4, 9, 10
ONS UK	~653,000 births/year (2025 provisional estimate)	Sections 4, 9
CDC US	~4.2M births/year	Sections 4, 9
American Academy of Pediatrics	60% of parents struggle with sleep; 30%+ infants have sleep problems	Sections 4, 10
Israeli PMC Study	37% of Israeli parents report baby sleep problems	Section 4
Israeli Sleep Ecology Study	72% of Israeli mothers suffer poor sleep	Section 4
Accenture Digital Health (2024)	77% of consumers open to AI-guided health advice	Section 4
Google Gemini API Pricing	Gemini 2.0 Flash: $0.10/1M input tokens, $0.40/1M output tokens	Sections 8, 9
Meta WhatsApp Business API	Per-message pricing ~$0.005-$0.05 (varies by type/country)	Sections 8, 9
Lebesgue / SuperAds (2025)	FB/IG CPM: UK ~$10.85, US ~$20.48	Sections 4, 6
PR Newswire — Huckleberry	Berry launch Feb 5, 2026; 5B data points; $14.99/mo or $119.99/yr	Section 5
ISOC-IL Israel Internet Index	98% WhatsApp usage among Israeli adults (2024)	Sections 4, 5
Statista	Smartphone penetration Israel: 91% (2025)	Section 4
World Bank	Israel fertility rate: 2.9 (highest in OECD)	Sections 4, 5

Appendix C: Consolidated Timeline — 36 Months

March 2026

Founders' agreement signed; Agentic AI Ltd. established; bank account opened

April 2026

Owner's loan ($100K) transferred; MVP development begins

June-July 2026

MVP ready; beta launch to 10+ users; sleep agent working on WhatsApp; AI accuracy >85%

August-September 2026

First paying customers (Dorit referrals); 25-50 customers; first revenue

December 2026 (Year 1 End)

100-300 customers (scenario-dependent); nutrition & siblings agents active; AI accuracy >90%

Q1-Q2 2027

UK market prep: English localization, GDPR compliance, multi-region deployment

Q3-Q4 2027

UK market entry; UK campaigns launch; first UK customers; 400-1,200 customers total

March 2028 (Year 2 End)

Operational break-even; loan repayment begins; 280+ active customers; MRR $7K+

2028

US market entry planning; VC round prep; product proven in 2 markets; 1,200-3,500 customers

End of 2028

Full loan repayment (base case: month 30-32); loan balance $0; full profitability

Appendix D: Detailed Competitor Comparison

Criterion	Numi	Huckleberry (Berry)	Smart Sleep Coach	Taking Cara Babies	Nanit	Israeli Consultants
Price	$400-$800/yr	$120-$180/yr	$60-$96/yr	$79-$179 one-time	$299-$399 + $149/yr	1,500-3,000 ILS one-time
Language	Hebrew native + English (Yr 2)	English only	English only	English only	English	Hebrew
Platform	WhatsApp + Web + FB + IG	App only	App only	Courses	Camera + app	Phone + WhatsApp (manual)
AI Type	Agentic — proactive	Reactive (Q&A)	Schedule-based	No AI	Computer vision + tracking	Human
Sibling Support	Yes — dedicated agent	No	No	No	No	Limited
24/7 Availability	Yes	Yes	Yes	No	Yes (monitoring)	No
Methodology	Dorit Kreizer — specific, proven	Multiple experts — generic	P&G research — generic	Cara courses — structured	Monitoring only	Varies
Long-term Memory	Yes — years, across children	No	No	No	Partial (camera data)	No
Business Model	Annual B2C + B2B for consultants	Monthly/annual sub	Monthly/annual sub	One-time purchase	Hardware + sub	One-time package

Key Insights:

Numi is the only competitor combining Agentic AI + Hebrew + WhatsApp
No competitor offers sibling logic (critical in 2.9 fertility rate market)
All global competitors are English-only — Israeli market completely open
Numi price higher than global ($400-$800 vs. $60-$180) but lower than private consultants (1,500-3,000 ILS), justified by WhatsApp, Hebrew, proactivity
Berry by Huckleberry is most serious global competitor, but remains reactive, English-only, no WhatsApp

Disclaimer

This business plan contains forward-looking statements based on current expectations and assumptions. Actual results may differ materially from projections. Financial projections are estimates and should not be considered guarantees of future performance. All market data is sourced from publicly available research reports and proprietary analysis. This document is confidential and intended solely for potential investors and strategic partners of Agentic AI.

Contact:
Agentic AI
Email: contact@getnic.ai
Website: getnic.ai

Document Version: 1.0 | February 2026 | Confidential