Skip to content

AI solutions architect · 0 to 1 builder · founder

I like problemsI have notsolved before.

and I want my hands on all of it

Taking a process apart and building it back is the part I actually enjoy. Strategy, design, the model layer, the funnel, the pricing, the campaigns. Six products of my own are live and selling, which is how I know the whole chain works and not just my favourite piece of it.

Idea to revenue, and every layer in between.

30+
AI products shipped
6
of my own, live and selling
60%+
margin per project
1
company built and sold

Six products. One year.

Every one of these shipped in 2025, and every one of them is live. Tap any of them.

yes, all in one year

The shelf

Six live products with real customers. On every one of them I did the strategy, the design, the LLM layer, the funnel and the growth engine that sells it. Some of them I ran end to end on my own.

01 / 06Legal-techbiroclick.co.il

BiroClick

Nine tools that get Israelis their money back, without a lawyer.

  • BiroClick. The claim starts free
    The claim starts free
  • BiroClick. Nine tools, one engine
    Nine tools, one engine
  • BiroClick. Analysis before payment
    Analysis before payment

The hard part

Nine claim types share one engine, but each has its own law, its own evidence and its own document. Making them one product instead of nine was the whole job.

Claim types
9
Analysis
Free
Document
9.90 to 29.90 NIS
Status
Live, selling
02 / 06E-commerceheroofyou.com

Hero of You

One photo becomes a kid who stays recognizable for twelve chapters.

  • Hero of You. One photo, one hero
    One photo, one hero
  • Hero of You. The likeness pipeline
    The likeness pipeline
  • Hero of You. Themes and tiers
    Themes and tiers
  • Hero of You. A story, chapter by chapter
    A story, chapter by chapter

The hard part

Generating one good illustration is easy. Keeping the same kid's face recognizable across twelve of them, so a parent says that is my kid and not that is close, is the entire product.

Chapters
12
Price tiers
14.90 to 42.90 USD
Ships to
33 countries
Status
Selling internationally
03 / 06Social appbambooil.app

Bamboo Connect

An 11,700-person community turned into something people actually meet through.

  • Bamboo Connect. Meet people near you
    Meet people near you
  • Bamboo Connect. Introductions by skill
    Introductions by skill
  • Bamboo Connect. Hosted meetups
    Hosted meetups
  • Bamboo Connect. Community tools
    Community tools

The hard part

A social app with no users is dead on arrival. This one launched on top of a community that already existed, so growth was never the problem. Landing an introduction at a moment when both people are genuinely free was.

Community
11,700 members
Since launch
800+ users in 2.5 months
Platforms
iOS and Android
04 / 06Vision AIfutureself-ai.com

FutureSelf AI

Idea to first paying customer in under three weeks.

  • FutureSelf AI. Your goals, rendered
    Your goals, rendered
  • FutureSelf AI. How a board is made
    How a board is made
  • FutureSelf AI. Your face in the scene
    Your face in the scene
  • FutureSelf AI. The tool surface
    The tool surface

The hard part

The interesting number is not in the product, it is on the clock. Idea to a stranger paying money in under three weeks, including the pipeline, the funnel, the pricing and the launch.

Idea to revenue
Under 3 weeks
Team
One person
Status
Live, paying customers
05 / 06Fitness SaaSfitil.app

FitIL

The growth engine was built before the audience existed.

  • FitIL. Training and management
    Training and management
  • FitIL. Programmes and workouts
    Programmes and workouts
  • FitIL. The exercise library
    The exercise library
  • FitIL. The trainer side
    The trainer side

The hard part

Three user types in one product, each of whom would happily be their own app: the trainee following a plan, the trainer writing it, and the gym owner counting both. Keeping that one coherent product is the design problem.

User roles
3
Status
Live, selling
06 / 06Empathic AIreliefai.life

ReliefAI

Here the prompt work is a safety problem, not a quality problem.

  • ReliefAI. A protocol, not a chatbot
    A protocol, not a chatbot
  • ReliefAI. An assessment tool
    An assessment tool
  • ReliefAI. The full tool set
    The full tool set
  • ReliefAI. The daily protocol
    The daily protocol

The hard part

Every other product on this page fails by being unhelpful. This one can fail by saying the wrong thing to someone in real distress, so tone and safety are the product, not a coat of polish on top of it.

Hardest constraint
Safety under distress
Model
Returning loop

The studio

The six above are mine. This is what I do for everyone else.

elya-studio.com

Elya Group

A studio that designs and builds tomorrow's products today.

Elya Group is the business behind the products. A boutique studio with a bench of 15 to 22 cross-functional specialists, structured in three arms so a client can come in at an idea, at an MVP, or at scale. Much of the pipeline arrives through the content engine built for it, which keeps compounding after each piece ships.

Visit the studio
Elya Group, elya-studio.com

Receipts

Four roles, one exit, and the through-line between them.

  1. Jan 2024 to now

    Elya Group

    Founder and CEO

    Three arms, so a client can come in at an idea, at an MVP, or at scale. Studio builds it, Vibe gets a validated idea to a working product in weeks, Strategy takes you from idea to PMF and fundraising-ready.

    • Shipped 30+ AI-native products: SaaS platforms, mobile apps, automation systems. Over 60% margin per project. Discovery, UX and UI in Figma, build and launch, with 15 to 22 contractors.
    • Built the prompting frameworks and dynamic context-loading that made the output better and the cost per run lower at the same time.
    • Built the group's inbound engine off 60+ long-form pieces engineered for AEO, GEO and search, instrumented in GA4, Search Console and Hotjar.
    • Consulting jobs left working AI workflows behind, actually adopted by marketing, R&D and analytics teams.
  2. Mar 2021 to Feb 2024

    Q-int

    Co-Founder and CEOAcquired
    • Owned product, ops and growth end to end, from positioning and roadmap through delivery, all the way to the acquisition.
    • Rebuilt how the team delivered and cut project turnaround by 40% in three months.
    • Ran the customer discovery that found the gaps behind a 30% retention lift in six months.
  3. Dec 2021 to Aug 2023

    Paycheck

    Product ManagerFreelance

    Alongside Q-int.

    • Owned feature definition end to end: requirements, specs and the design plan, prioritized against business goals.
    • Drove the cycle from spec to release to actual adoption inside the client's org.
    • Shaped the pricing and packaging logic and iterated go-to-market on analytics.
  4. Jan 2020 to Sep 2020

    Palto

    Co-Founder and CPO
    • Ran 100 discovery interviews with practicing lawyers and turned them into the full product definition for a legal-tech platform.
    • Owned UX and feature prioritization, and rebuilt onboarding and activation.
    • It wound down after the Israeli Bar Association challenged the model on regulatory grounds. A market constraint, not a product one.

What actually moves me

Every one of these is a thing I like doing, next to the thing that proves I have done it. In that order, because anyone can claim the first half.

this is the real pitch
  • New markets, not one lane

    Six products in six categories: legal-tech, e-commerce, a social app, fitness SaaS, vision AI and an empathic-AI health tool. Each one meant learning a different customer from scratch.

  • Taking a process apart

    BiroClick runs nine kinds of legal claim on one engine. Each has its own law, its own evidence and its own document. Making that one product instead of nine was the entire job.

  • Ceilings people say are fixed

    Bamboo Connect reached 800+ users in two and a half months from launch. Q-int, which I co-founded and ran as CEO, was acquired.

  • My hands on the whole thing

    I do not take work where I only touch a quarter of it. On my own products that means the strategy, the design, the model layer, the funnel, the pricing and the ads that sell it.

  • People, mostly

    The reason any of this is worth doing. I am always looking for new people to build something good with, which is the whole point of this page.

What I actually do

AI and automation

  • LLM integration via APIs
  • Prompt engineering
  • Context management
  • Agents and workflows
  • Process automation
  • Make (Integromat)
  • AI-assisted delivery

Product

  • 0 to 1 and 1 to n
  • MVP scoping
  • Roadmaps
  • UX flows
  • Specs, QA and docs
  • Design thinking
  • Pricing and positioning
  • CRO

Growth

  • AEO and GEO
  • SEO
  • Funnels
  • Conversion optimization
  • Google, Meta and LinkedIn ads
  • Pinterest and Reddit ads
  • GA4
  • Search Console
  • Hotjar and Clarity

Tools

Leading the work

  • Vendor and contractor management
  • Stakeholder management
  • Deciding without full information
  • Coaching and high ownership

Been there

  • AI
  • SaaS
  • B2B
  • B2C
  • E-commerce
  • Finance
  • Legal-tech
  • Real estate
  • Apps

Studied

  • 2015Business Administration and ManagementThe Open University of Israel
  • 2019Innovation and Startup ProgramsJolt, Tel Aviv
  • 2023Generative AI Tools for Business LeadersLinkedIn Learning

Also doing

Yad Labanim

Steering Committee, Bereaved Brothers

Tel Aviv, Dec 2025 to now

Selected from thousands of bereaved brothers to lead the organization's national plan for securing rights and building programs for bereaved siblings.

Tell me what you are building.

I read everything myself. If it is a fit I will say so fast, and if it is not, I will say that too.