Roméo Roy

My projects

Everything I've launched. With the numbers, zeros included.

Not a showcase. An honest inventory: what's alive, what's asleep, what's starting, and what it earns. All of my own numbers. My clients' and partners' stay anonymous.

Last updated: September 15, 2026. The full numbers are redone every month, on video.

  • Live since 2023

    JePréviens

    Mobile app · neighborhood alerts

    30,000
    users
    TF1 + M6
    prime-time national TV
    Town halls
    equipped across France, paid dashboard

    A free app that lets the residents of a town warn each other. Built solo, with AI blurring and AI moderation running in production. It aired on the 8pm national news: 10,000 sign-ups in one minute, and a €400 Google Maps bill the same evening. Two years at close to zero revenue. Since summer 2026 town halls pay for the setup (€490) and €199 a year above 2,000 residents: towns equipped in several regions, one that paid for a custom feature, others waiting on a council vote, and town halls now reaching out on their own. It's still nowhere near paying for the time I put in, but for the first time money is coming in.

    A free product with 30,000 users is not a business. It's proof, and it's a school.
  • Live since 2023

    Smoke-Defender

    E-commerce · anti-burglary smoke generators

    300+
    customers
    ≈ €2,400
    of monthly revenue (summer 2026)
    35
    countries shipped to since September 2026

    My own e-commerce business, A to Z: sourcing, a store built with AI, Meta and Google ads I've been running myself for three years, sales on Amazon, and the very real customer support that comes with it. I'm also part of the manufacturer's development team: electronics, hardware, packaging, we adapt the product together from what my customers report. It's the steadiest revenue line in the portfolio, and the one a better-funded competitor could take from me. I know that, and I say it. So I stopped holding it back: since September 2026 I ship across Europe, ads run in France, Italy and Germany, and I publish the numbers every month.

    Revenue that depends on a product you don't manufacture can be caught up with. All the more reason to push it hard while it's moving, and to measure it monthly.
  • Live since summer 2026

    Training

    Training · 3 days on your real project

    2
    contracts signed this summer
    3 days
    the format I measured
    €0
    out of the first trainee's pocket

    The work I enjoy most, and the one that pays best, with no after-sales support. First trainee: three days, her site live at the end, fully funded. Then the training organisation that handled her paperwork bought the same training from me to rebuild its own website. This is what this site sells, and what the YouTube channel sells.

    The site gets built in a day and a half. The value is in the next day and a half: actually going live, and being found.
  • Live since relaunched September 10, 2026

    YouTube channel “Roméo Roy”

    Media · one video a week, in French

    382
    subscribers at relaunch
    22
    videos over several years, before that
    1 / week
    the commitment, from now on

    The old “SOS Développeur” channel, renamed. My struggles, my wins, and my revenue project by project, every month. The one thing I've never managed to hold is consistency. And it happens to be the only price to pay for building something nobody can take away from me. Four days after the first video, a stranger wrote to ask how I'd found a client on Reddit. The answer will be a video.

    An audience under your own name is the only asset a competitor, a client or a partner cannot take back.
  • Live since 2025

    Lean Intelligence

    SaaS · AI content generation for a US agency

    200+
    active users
    1 / month
    “Talk with an AI expert” show
    Reddit
    where I met the client

    A social-media content generation tool built for a marketing agency in the United States. The client found me on Reddit. On top of the product, I host their monthly AI show, in English, from my desk. Proof that you can invoice in dollars from the French countryside.

    A small project delivered well turns into a long relationship. Every good client relationship I have started small.
  • Live since 2026

    Axiotrans

    SaaS · route management for road carriers · co-founder

    2
    app stores (iOS and Android)
    50 / 50
    with my co-founder
    €0
    collected, first prospects on the phone

    The only project where I'm not alone: I co-founded it with the owner of a transport company I met online. He knows the trade and the field, I build the product. It grew out of a custom app I had built for a regional carrier: delivering it, I realised every other carrier had the same problem. Fully working app, live on both stores, payment in place, and since September 2026 the first prospects are calling us. Zero euros collected to date, and I'll say so the day that changes.

    Every job you repeat for a client hides a product. But a product with nobody to sell it stays an app nobody installs.
  • Being built since 2026

    PostIt

    App · with my American colleague

    2
    people, two continents
    €0
    of revenue

    An app I'm building with the American colleague I met through Lean Intelligence. Too early to say more, and above all too early to promise anything. I'll talk about it in the monthly numbers once there are numbers.

    I plant seeds everywhere. The problem is when they all grow at once.
  • Live since 2025

    n8nEye

    Mobile app · monitoring for n8n automations

    2
    app stores (iOS and Android)
    $271
    in sales, worldwide
    $0
    of advertising

    n8n is the tool I build my automations with, for myself and for clients. The problem when you have a lot of them running: you don't know one has broken until someone complains. n8nEye puts all of that in your phone: the state of your workflows, failed executions, an alert when something falls over. I built it for myself, then published it on both stores, in English, with a website and a solid store listing. No advertising, ever. There was only one competing app, badly done: I rebuilt it better. Result: sales all over the world, $271 to date. That's not revenue, it's a market test that worked.

    A need you have yourself, one badly-built competitor, a good store listing: that's enough to sell with no ads. And $271 mostly tells me where people pay for an app: outside France.
  • Being built since January 2025 · rebuild started September 15, 2026

    1Bleu

    Mobile app · site notebook for the trades · rebuild underway

    20 months
    without a line of code, and still downloaded
    €0
    of revenue, free since day one
    €9.90
    a month, the Pro version to come

    A visit-notes notebook for tradespeople: a photo, a voice memo, the client, the place, and it's filed. Published on both stores in early 2025, then left alone. Twenty months later it's still being downloaded with zero promotion. I don't even know the exact number of users: that will be in the next monthly report. A demand that persists while nobody tends to it is rare, so I'm rebuilding it entirely, with AI to transcribe and summarise the notes, and a Pro subscription for those who want more. Six-month goal: 100 subscribers. Otherwise it goes back on the shelf, and I'll say so.

    An abandoned app that keeps getting installed is a need that's waiting. It's not a business yet: it's 100 subscribers or nothing.

And for clients

What I've built to order.

Apps that run, for people who have no IT department. I don't name them: in a village, everyone recognises everyone.

  • A regional road carrier

    Web and mobile app for route management, delivery slips, signatures. In production. The origin of Axiotrans.

  • A building contractor

    Tablet app for site foremen: measurements, on-the-spot quotes, crew tracking. Delivered.

  • A work-integration workshop

    One platform for two operations: green-space crews in the field, and refurbished equipment resold online. In production.

  • A rural guesthouse

    AI-assisted search ranking, automated publishing, monthly reporting. The model for light recurring work that suits a very small player.

  • My own town hall

    The village website, back when I was a municipal councillor. A simpler version, built entirely with AI, is under discussion with the new team.

Want to do the same, at your own scale?

Most of these projects, you can learn to build the first version yourself. That's exactly what I train people to do, on their own project.

Book a free 15-min call →

Free, no commitment. Anything you book afterwards is paid at booking.