Nutribuddy

Inspiration

I work with food labels pretty often, so I just wanted to make a tool that made that process a little easier .

What it does

Work together with an AI agent to turn recipes into nutrition label.

How we built it

Call data from USDA api, used a lighter weight gemini-2.0-flash to handle chat requests, and used langgraph to create a framework/give tools for the chatbot to actually make the nutritiion labels/ interact with the user.

Challenges we ran into

Figuring out things inside the GC console was confusing at first.

Accomplishments that we're proud of

Creating a real app that does something.

What we learned

I used this to learn more about FastAPI and SOLID architectural principles and integrating an LLM into a real application.

What's next for Nutribuddy

Possibly add a graph-based recommendation system for like certain ingredients that go together, and add more formats for the nutrition labels.

Built With

• gcp
• langgraph
• python
• react