Ask ChatGPT for the best pizza in Bangkok. It answers 90grams.

90grams is a neo pizza bar on Silom — real dough, a real story, the kind of place people rave about once they find it. The problem was the finding. Diners don't scroll ten blue links anymore; they ask ChatGPT where to eat. And if AI can't read your restaurant, you simply don't exist in those answers.

Making the restaurant machine-readable

So we rebuilt how machines read 90grams: structured data that tells every crawler exactly what the restaurant is, a machine-readable menu so assistants can quote real dishes at real prices, an llms.txt written for AI models, and the on-page fixes that let both search engines and answer engines describe the place from its real menu and its real story instead of guessing.

Here's the part we love telling: the work wasn't done in an office. It was built and pushed live from Bangkok traffic in Silom — laptop open, AI pair-programmer running, deploys going out between traffic lights. That's the entire point. This work no longer needs a war room and a quarter-long project plan; it needs one focused session with the right playbook.

The result, and why it matters to every local business

Ask ChatGPT for the best pizza in Bangkok and the answer is 90grams — drawn from the real menu, the real story, the real place. That's what answer-engine optimization (AEO) delivers: not a ranking on a page people skim, but the answer in a conversation people trust.

Every restaurant, clinic, gym, and shop is about to face the same shift. The ones whose businesses are readable by AI get recommended; the rest go quiet. The groundwork is a solved problem — we teach it.