inputs
taste profile location time window mood budget effort level
Project / TasteGraph
Recommendations should understand the moment.
TasteGraph is a food decision engine for mapping taste, context, constraints, timing, craving, and usefulness into better recommendations.
Problem
Most recommendation systems flatten the reason behind the choice.
“Best restaurant” is usually the wrong question. The better question is: what fits the person, place, mood, timing, effort level, and constraint set right now?
System shape
Taste as structured decision context.
scoring
fit novelty comfort friction confidence timing
output
recommendation why it fits tradeoffs backup option next signal
Product direction
A decision engine before a directory.
Personal recommendation layer
Recommendations shaped around actual preference memory, not generic ratings.
Context-aware discovery
Good for choosing food while traveling, hosting, commuting, or trying to avoid decision fatigue.
Reusable taste graph
The same structure can support restaurants, recipes, cocktails, playlists, and creative pairing systems.
Constraint
Do not build another ratings site.
The leverage is in reasoning about fit, not collecting another list of places. TasteGraph should explain the recommendation clearly enough that the user can correct it.