No Employees, Seven AI Products, $1M a Month: How Thibault Louis-Lucas Builds First and Chooses Later

From bankruptcy to $1 million a month

French founder Thibault Louis-Lucas, known online as Thibault Maker, now runs seven AI products that together bring in more than $1 million a month. He has no formal employees and has raised no investment. A column by Korean entrepreneur Lee Jong-dae, republished by Mobiinside, traces the journey using only figures Louis-Lucas has shared publicly. The one-line summary: he stopped choosing what to build and started building to find out. It’s an extreme case of fast product validation.

• Revid.ai turns text, links or voice into vertical videos for TikTok, Reels and Shorts — about $600,000 a month.
• Outrank.so uses AI to decide what content will rank and generates it — about $300,000 a month across more than 2,500 sites.
• Others include Feather (Notion to blog), SuperX (X growth) and PostSyncer (multi-platform publishing).

His first two companies, founded in 2015 and 2017, followed the textbook: raise more, hire more. Paying customers never materialized, salaries went unpaid and the business was wound up in court.

Principle 1: Don’t pick — build and let the market choose

The usual path is to research, pick one idea and invest months in it, even though no one knows the answer at the picking stage. Louis-Lucas reversed the order. AI made that possible: most of his products are thin wrappers around existing models, so each can ship in weeks.

Revid.ai was his tenth attempt after nine failures. It started as Typeframes, a tool for simple text animations; when users kept making short vertical videos, he pivoted and renamed it in September 2024. Follow user behavior, not the original plan.

Principle 2: Charge from day one — and don’t charge too little

“Charge from day one. That’s how you know if people really want it,” he says. Free launches blur real demand. His price range is $50 to $100 a month, which filters out customers who churn quickly. For a solo operator, a few customers paying properly beats many paying little.

Distribution comes from search, not ads: he builds free tools that match real search phrases, such as “voice to video,” and converts satisfied users to paid products. Outrank productized that tactic and grew from $300 a month to $300,000 a month in 15 months.

Principle 3: Don’t go all-in on the winner

Even after Revid took off, he kept launching. His goal: five SaaS products at $100,000 a month each. Scaling one product requires a team — the structure that sank him twice — and AI tools built on others’ models and platforms can be shaken by a single policy change. In June 2024 he bought Feather for $250,000 when it made $7,000 a month and doubled it to $20,000 within six months.

In 2022 he sold Tweet Hunter and Taplio to lempire for a final $8 million, but only $2 million came upfront; $6 million depended on hitting revenue targets over two years, a deal he now publicly regrets. His September 2025 update showed two of five products past target, two growing and one — Feather — marked as a miss. “Honestly, building SaaS is still hard,” he wrote.

Which problem are you on?

1. Don’t know what to build? Ship something instead of deliberating.
2. Don’t know if it will work? Charge from the start, and not too cheaply.
3. One thing worked — now what? Don’t bet everything; stop at a size you can handle and add the next.

There are conditions: he can code, he builds card-paid, sales-free, lightly regulated tools, and his exit gave him runway. The transferable lesson is the method — turn unknowns into things you can verify quickly. For contrast, see David Holz, who rejected Apple twice and bet everything on one product, and a one-person company sold for $80 million.