Build notes

Two ideas.
No filler.

What I am learning while building Meet My Menu.

A readable menu is not the finish line

The goal is not extracting words. It is helping someone choose a meal without handing the decision to someone else.

That changes what I build. Camera guidance matters. Clear categories matter. Honest allergen warnings matter. Recovery after a bad scan matters.

The test: can someone get from a menu to a confident choice on their own?

The hardest part happens before the scan

A camera can read a menu only after the person using it knows where to aim.

“Move left” is more useful than “try again.” Good guidance should say what changed, what to do next, and when the full menu is in frame.

The rule: every error should help the next attempt succeed.

A menu decides whether a restaurant gets considered.

Research behind Meet My Menu: 85% of diners look at a menu online before choosing somewhere new.

An accessible, structured menu lets people make that choice independently and gives search and AI systems real information to work with.

Read the full menu research on Meet My Menu ↗

AI is becoming a restaurant's new front door.

About one in five diners has used AI to choose a restaurant, where answers surface only a small shortlist.

Machine-readable menus help restaurants show up with accurate details when diners ask voice assistants and AI tools what to eat.

Read the AI restaurant discovery research ↗

That is the whole blog for now.

Have something worth discussing? Let’s talk