Inspiration
Most healthy-eating apps ask you to give up the food you grew up with. I wanted the opposite: eat healthier while keeping the flavors and culture you love. In the Dominican Republic, food is identity — mangú, rice, plantain — so a plan that ignores that never lasts.
What it does
Reset Plate is an AI agent. You type a few things you love — dishes, restaurants, singers, movies — and your city. The agent:
- Resolves each taste into Qloo entities
- Expands those tastes through Qloo's Insights API into restaurants in your city
- Filters them with simple, sustainable "Reset" rules (one balanced plate, kitchen closes at night, a 10-minute walk after the biggest meal)
- Plans a 7-day week: one matched, healthy-leaning restaurant per day, two home-cooked days, a plate tip for each, and an explanation of why every pick was chosen
How we built it
Python and Flask, calling Qloo's Search and Insights APIs. It runs as a service on an Ubuntu server (Hetzner). A scoring layer ranks Qloo's results toward healthier options without throwing away the user's taste.
Challenges
Mapping very local tastes (Dominican dishes and artists) to Qloo entities, and balancing "healthy" against "what you'll actually enjoy" so the plan feels personal, not like a diet.
What we learned
Taste is a stronger lever for habit change than rules alone. Qloo's cross-domain graph lets music and movie tastes inform food choices.
What's next
Recipes for the home-cooked days, a Spanish version, and a chat/voice interface.
Log in or sign up for Devpost to join the conversation.