How GastroMatch works
GastroMatch helps users reduce uncertainty when choosing a dish in an unfamiliar context.
The problem
- Restaurant ratings do not help with choosing a specific dish.
- Menu descriptions are incomplete.
- Translation is not enough.
- People need a worthy choice they can trust.
The decision model
- GastroMatch works at dish level, not restaurant level.
- It acts as a personal decision engine.
- The goal is not perfect prediction, but trusted guidance.
Recognize
- The user can photograph a dish.
- The system identifies it.
- It explains what it is.
Interpret
- The system explains ingredients, context, and why the dish may fit the user.
- This is not just recognition, but interpretation.
Remember
- The user saves the dish, rating, and place.
- This strengthens Taste DNA.
- Future recommendations become more useful.
Two confidence sources
- Local dish signal — from people with similar taste nearby.
- Personal mode — based on Taste DNA alone.
- Both are valid; the user chooses what to trust.
Trust over coverage
- The system does not need to answer every time.
- Silence is better than a confident error.
- Explainability matters more than aggressive recommendation.
Try it
Open the app and try your first dish or menu scan.