Inspiration
We wanted to tackle food waste in a fun, engaging way. Instead of letting leftovers go to waste, what if cooking became a game where you can level up with friends, all while helping the planet?
What it does
Crumb&GetIt lets users scan their fridge or type ingredients, then instantly generates a recipe. Users can chat with celebrity chef AI, follow step-by-step videos, and earn sustainability points by completing dishes.
How we built it
We used React for the frontend, Flask for the backend, Gemini APIs for recipe generation and validation, MongoDB Atlas for storage, and integrated text-to-speech for an immersive AI chef conversation experience.
Challenges we ran into
Fine-tuning AI outputs to generate clean, structured recipes. Making AI conversations feel natural and exciting, like talking to Gordon Ramsay. Ensuring accurate recognition of fridge ingredients from pictures. Seamlessly tying together multiple APIs (Gemini, MongoDB, speech synthesis) while keeping the user experience smooth.
Accomplishments that we're proud of
Built a working multimodal AI flow (text and image to recipe). Made cooking feel fun, personal, and rewarding with live chats and point systems. Created an end-to-end working app with clean UI, real-time database integration, and live scoring.
What we learned
How to design intuitive flows combining AI, gamification, and social competition. How to integrate and orchestrate multiple AI services in real-time. The importance of feedback loops (e.g., scoring and chatting) to keep users engaged.
What's next for Crumb&GetIt
Integrate grocery store APIs like Melissa to help users shop missing ingredients nearby. Expand celebrity AI chefs users can interact with. Launch full mobile support. Refine recipe scoring with even smarter sustainability analysis.
Log in or sign up for Devpost to join the conversation.