Skip to content
Jonathan Naal

Product Manager & Applied AI · Paris, France · Open to relocating

I advise, design and build AI systems that turn complex business workflows into simpler, scalable products.

Let’s keep it simple: I love building with AI, deep thinking and understanding how things work. On the business side: venture building, B2B go-to-market, and turning an edge into revenue. Focus on outcomes, fast prototyping, and owning products people actually use.

Connect on LinkedIn (opens in a new tab)Follow on GitHub (opens in a new tab)See projects
Pixel-art portrait of Jonathan Naal: curly brown hair, round gold glasses, a black blazer over a white shirt.

Work

What I can build for you

  • Applied AI

    Build AI systems

    Agents and Claude Skills wired into your tools, with guardrails built in.

    Examples · gtm-vision-pnl (opens in a new tab), Plania

  • Product

    Design products

    From the first user interview to a demo, then an app used every day.

    Example · casa-liva (opens in a new tab)

  • Strategy · Consulting

    Solve business problems

    Discovery calls to find the AI use cases worth building, then onboard your team.

    Examples · CNMSS, LGM

  • Business · Growth

    Grow the venture

    Business development, go-to-market operations and growth, leading the people, budget and time it takes.

    Example · Plania

How I work

From the first discovery call to daily use.

  1. 01

    Systems thinking

    Map the full workflow

  2. 02

    Customer-first

    Users shape the draft

  3. 03

    Ship & iterate

    Build, test, improve

    Iterate from step 03 back to step 02 · or stop, and write down why

  4. 04

    Ownership

    Stay until it’s adopted

Path

Where I came from

Engineering taught me the theory and how to break problems down. Consulting taught me to explain them and pick the right ones to solve. A start-up taught me to ship, in tech and in business.

Education 2020 – 2026

  1. Pixel-art Jonathan in a PSI* hoodie writing in a notebook at a desk with a stack of textbooks, a PSI* mug, a calculator and a laptop under a desk lamp, in front of a whiteboard of formulas with a postcard of Valbonne, beside a crescent moon.

    2020 – 2023

    Preparatory classes, PSI*

    Centre International de Valbonne · France

    Long nights of maths, physics and engineering sciences in a starred class, the top level of France’s preparatory classes, aiming for the top-tier engineering schools.

  2. Pixel-art Jonathan in an EPF hoodie presenting his thesis on steering a digital product with AI agents to classmates, pointing at a whiteboard of diagrams, with the Eiffel Tower at sunset through the window.

    2023 – 2026

    Engineering degree, MSc level

    EPF Paris · France

    A generalist engineering degree, major in Engineering Management and Data/AI.

    • Experience during school: CNMSS, LGM Industrial Engineering Group and TechGuys Consulting.
  3. Pixel-art Jonathan in a UCSB hoodie working with classmates around a table of laptops, notebooks and business books, in front of a whiteboard reading “From Data to Impact”, with palm trees, the campus tower and a sunset over the ocean behind them.

    2024 – 2025

    UIP program

    UC Santa Barbara · California, USA

    Business American Dream: I followed the Technology Management Program (TMP), with business strategy, marketing and data analytics taught by exceptional professors.

    • My best career decision: engineering thinking, now paired with business thinking.
  4. Pixel-art Jonathan in a suit presenting his thesis to a jury, pointing at a slide on the shared operational memory, the controlled action surface, the steering loop and where agents belong.

    2026

    Steering a digital product with AI agents

    EPF Paris · Master’s thesis

    Designed while running Plania: a management system that lets a small team understand, act on and steer a product with AI agents.

    • Three building blocks: a shared operational memory that makes the product legible to humans and AI, a controlled action surface connecting agents to its tools, and a steering loop from goals to learnings.
    • Where agents belong: where no predefined path gives the answer and unknowns must change the tool’s behaviour. Deterministic tools for the known, like a V-model; agents for the unknown, like agile.

Professional experience 2025 – 2026

  1. Pixel-art Jonathan speaking into a mic at a lectern before a military audience, beside a slide on AI strategy and transformation: assess, prioritise, experiment, deploy.

    Jun – Aug 2025

    AI Strategy & Transformation Consultant

    CNMSS · French military health insurance · France

    Prepared and fostered AI adoption: choosing with leadership where AI was worth deploying, then bringing the teams along.

    • Evaluated a vendor’s AI proof of concept from the buyer’s side and prioritised the use cases worth building.
    • Gave an AI awareness talk for non-specialists, leadership included, then ran adoption workshops.
  2. Pixel-art Jonathan and his team presenting an AI roadmap for LGM’s technical documentation in a meeting room, from the current process to a phased implementation plan.

    Sep – Dec 2025

    Digital Transformation & AI Consultant

    LGM Industrial Engineering Group · France

    Led an international team defining an AI roadmap to streamline the technical documentation LGM sells to B2B clients.

  3. Pixel-art Jonathan holding a laptop and pointing at a screen on AI-powered product delivery and the team’s adoption of Claude Code and ElevenLabs, in an office overlooking a city at night.

    Feb – Aug 2026

    Product Manager

    TechGuys Consulting · Canada

    From Scrum Master to Product Manager: delivery first, then B2B client accounts from discovery to deployment.

    • Designed AI-powered delivery workflows wired into the company’s tools, drove their adoption across the organisation and made them the internal standard.
    • For one client, led the design and deployment of an AI browsing and sales agent with its sales leadership.
    • As my role shifted toward Plania, drove Claude Code and ElevenLabs adoption across teams.
  4. Pixel-art Jonathan at his desk in the Plania office, facing screens of running AI agents, the product site, Claude Code and a database diagram, under a whiteboard sketch of the AI-agent-first system.

    May – Sep 2026

    Venture Lead

    Plania · TechGuys Consulting · Canada

    Took the lead of Plania, an AI SaaS for business plans and financial projections, from product to growth.

    • Built an AI-agent-first system, the core of my thesis: agents run marketing, product, the website, customer support and finance, with a human approving what matters.
    • Led the shift from SaaS to marketplace, launched new subscription offers and grew the user base through acquisition, partnerships and community.

Entrepreneurial projects

Above all, these projects were a massive learning experience. Some stayed early-stage, others brought in some revenue, but the aim was to learn and run the whole loop: idea → solution → adoption by a user → outcome, without necessarily scaling it afterwards.

  • E-commerce

    Online store

    My first online business, and my first view of how a whole business fits together, from the store to acquisition and orders.

    Shopify · Google Ads · DSers

  • Social media

    Instagram management & monetisation

    Managed and monetised Instagram accounts, then built the team and the operation behind them.

    Acquisition channels · Content strategy · Team recruiting · Workflow design · Internal training · Automation · Contracts · Meta tools

  • No-code

    SparkApp

    A motivation app for aspiring entrepreneurs, and my first no-code build.

  • AI SaaS

    Max Agent

    An AI SaaS for real-estate agents, built with a team of three I put together.

  • Operations

    Brain OS / Taskforce

    Started as I moved from Product Manager to Venture Lead at Plania, out of the many problems my right hand and I solved along the way.

  • Data · Private equity

    Helm

    An AI-powered value-creation solution for private equity funds after an acquisition, helping operating partners improve EBITDA.

Want to know more?

Today

Applied AI engineer, building solutions and growing my business skills.

What comes next? Maybe you.

Let’s shake on it

Personal projects

Built in public

I build solutions for my daily life with AI agents, and I try a lot of them. Every new feature or technology sends me exploring until I find it a concrete use. Today most of them are powered by Claude, wired to tools and services through APIs.

Featured · TypeScript · Live evals

Consultant Dots

A practice partner for tech-consulting case interviews.

Consultant Dots structures a business case the way a consultant would. A fixed Claude pipeline separates facts from assumptions, checks every quote against the case and waits for your answers before proposing options, a roadmap and an oral pitch. An agentic interview mode then plays the client, so you can practise what AI can’t do for you: asking the right questions and defending a recommendation. It runs locally with Claude Code or the Claude API.

  • TypeScript
  • Next.js
  • Zod
  • Claude Code
  • Anthropic SDK
GitHub: Consultant Dots (opens in a new tab)

  • JavaScript · 16 stages

    open-ats

    See your application the way a screening system does.

    open-ats runs a CV and a job posting through a 16-stage recruiting pipeline, then returns a score across four filters, a verdict computed in code and a rewrite that only claims what the CV can prove. It works inside your Claude Code session, with no API key required, and keeps every file on your own disk.

    • JavaScript
    • Node.js
    • Express
    • Zod
    • Claude Code
    GitHub: open-ats (opens in a new tab)

  • Python · Human in the loop

    GTM Vision P&L

    Outbound prospecting, with a human approving every send.

    GTM Vision P&L is a Claude-first workflow that runs the design-partner search for Vision P&L, an early-stage product built to show Shopify merchants the real margin of every order. Claude Code and two Skills enrich accounts through Clay, identify the decision-maker and draft a co-branded e-mail; nothing leaves Gmail without explicit approval, and a meeting is recorded only once Calendly proves it. The code is public for viewing only.

    • Python
    • Claude Code Skills
    • Clay
    • Gmail
    • Calendly
    GitHub: GTM Vision P&L (opens in a new tab)

  • TypeScript · Used every day

    Casa Liva

    A shared family calendar that finds time to spend together.

    Casa Liva is a family planner shaped by its users’ requests. Casa AI, powered by the Anthropic API, reads the household’s calendars, including shared Google Calendars, finds a free slot and drafts the event, which is created only once someone confirms it. ElevenLabs adds voice dictation and spoken briefings. Invite-only for now, and in daily use by one family.

    • Next.js
    • Supabase
    • Anthropic SDK
    • ElevenLabs
    • Google Calendar
    GitHub: Casa Liva (opens in a new tab)

  • TypeScript · Live · Free

    Signature Studio

    Animated e-mail signatures, made in your browser.

    Signature Studio turns a portrait into a looping black-and-white GIF signature for Gmail, Outlook or Apple Mail. Place two guides on your face, add your details and paste the result. It is free and needs no account: the original photo never leaves the browser, and publishing uploads only the finished GIFs, to a public link.

    • TypeScript
    • Next.js
    • Vercel Blob
    • Vitest
    GitHub: Signature Studio (opens in a new tab)Live app: Signature Studio (opens in a new tab)

Currently

Pushing the limits of creativity every day with ever more ambitious projects. I use AI to help me get there and to feed my curiosity, from software to hardware. Have a project in mind you’d like Claude to run? Walk me through it: I’ll come back with a working demo.

Pixel-art Jonathan in headphones at his desk, working on a laptop beside a screen of Claude running a web app, a data analysis, a marketing campaign and a hardware prototype, with a small drone and a 3D printer nearby.

Contact

Let’s talk

Write to me by e-mail or on LinkedIn. Tell me the workflow, who runs it today and what a good outcome looks like.

Jonathan Naal

Product Manager & Applied AI

Paris, France

E-mail jonathannaal.official@gmail.comLinkedIn in/jonathannaal (opens in a new tab)GitHub spykernv (opens in a new tab)
Pixel-art Jonathan in a suit and headphones, laptop under his arm, walking to a plane at dusk, with luggage tagged CDG, YUL and LAX and the skylines of Paris, Montréal and California behind him.