What we learn building products people actually find.
Technical notes on product development, SEO and visibility inside AI answer engines. Written in Geneva.
AI Agent in Production: What Breaks First
An AI agent that passes every test can still break in production for five specific reasons, almost never tied to the model itself. The checklist to catch them before launch.
GPTBot, ClaudeBot, PerplexityBot: Block or Allow?
GPTBot protects training, OAI-SearchBot feeds ChatGPT answers: mixing them up loses visibility without gaining protection. Here is how to decide, crawler by crawler, with a working robots.txt configuration.
llms.txt: a practical guide and its real limits in 2026
A well made llms.txt takes an hour to write. Here is how to structure one, a real example, and the 2026 data that honestly shows what it changes and what it does not.
GEO: getting cited by ChatGPT, Perplexity and AI Overviews
A share of searches no longer ends in a click. Here is how answer engines pick their sources, the six levers worth pulling, and how to measure progress without paid tooling.
Forward Deployed Engineer: the engineer who replaces the consultant
An engineer who starts from the problem rather than the ticket, and ships software to production rather than a recommendation. What the role covers, and when it is the wrong choice.
Shipping a product in 90 days: what AI changes, and what it does not
The figure sounds impressive, so it deserves taking apart. What AI genuinely accelerates in a project, what it leaves untouched, and the three client-side conditions for holding the pace.