Apps I ship
Ship activity — production deploys, live
117 of 136 days shipped something
Here is a look at my personal portfolio — 14 products shipped since Feb 2025 on my stack of choice (Next.js, Supabase, Vercel, Claude Code as the AI layer), working toward 30 in 2026. Below are 4 selected.
Every venture below is run day to day by AI employees — the Submarine system — with me as the captain. Honest statuses, no vanity metrics.
live paying · launching building now · pilot testing · autopilot running unattended · active ongoing engagement
ItsMyLive
Live-music booking marketplace connecting artists with restaurants, hotels and bars in France. Co-founded.
Water Taxi Boat
The first marketplace for boat taxis on the French Riviera — paying operators onboard.
Water Taxi OS
SaaS back-office for boat operators — requests, quotes, payments, availability.
Daily Boat Rental
Boat-rental brokerage on the Côte d'Azur — recurring revenue.
Book a Boat Doha
Boat charter brokerage in Qatar — same engine, new market.
Beyond Riviera
Curated experiences brokerage on the Riviera.
Coworking Doha
Flex-office brokerage in Doha — introduction-fee model.
WithClaude
AI employees deployed into SMEs, one workflow at a time — the productized version of how I run everything else.
GCC Deep-Tech Atlas
Data app mapping deep-tech ventures across the Gulf, built on a multi-signal sourcing radar.
FleetBridge
B2B arbitrage marketplace between car-rental companies — fully anonymized supply and demand.
Lighthouse Business Advisory
Innovation-funding and market-entry advisory between France and the Gulf (QFC, Doha).
The SEO factory
A portfolio of local lead-gen services — events, drone shows, event furniture, facade renovation, kitchen-hood install and cleaning, grants content — run end-to-end by AI employees.
Qui paie ?
Addition Roulette — my first B2C consumer app, a bill-splitting game for group dinners.
Delivery to Yacht
Provisioning delivery to yachts at anchor — drone drop and boat shuttle, commission on order value.
I don't have a good idea rate. I have a good kill rate. Every venture above passed a playbook I run on everything before it gets a repo:
Candidates come from adjacent markets, existing assets, or competitors already proving demand — never a whiteboard.
A zero-cost, 90-minute pass against hard gates: real buyer, real margin, a market that already exists. Most ideas die right here.
Survivors get a real signal before any code — a smoke-test landing page, cold calls, a paid ad.
I model the actual €/month at month 12. If the math doesn't clear the bar, it doesn't ship.
Only what survives all four gates gets a repo.
69 ideas run through this filter so far · 32 killed at the desk, before a line of code
Not every idea survives contact with the market. Here is a short, honest log of a few I killed on purpose — and why.