Inspiration
We’ve all been there—cooking at home to save money, and there are a lot of us who'll definitely agree with me on "Oh I'm done with McDonald's or Chipotle" That’s how the idea for PeerPlates was born: connecting people who crave wholesome, home-cooked meals with fellow students who love to cook. Why not match the two?
What it does
PeerPlates is an AI-powered app that pairs home chefs with students looking for nutritious, home-cooked meals. Users can search for meals based on their preferences, and the AI recommends the perfect dishes, making meal-sharing fun and effortless.
How we built it
Users can describe their cravings in plain English, like "I want something spicy but high in protein," and PeerPlates’ AI, powered by Azure OpenAI, translates this into meal recommendations. We also use Azure Computer Vision to analyze images and extract nutritional data from user uploads. The backend, built with Python's FastAPI, orchestrates these services and manages image storage through Azure Blob. On the frontend, we built a user-friendly interface using React.
Challenges we ran into
One major challenge we an implementation flaw which we realized only during a dry run. With only 5-7 documents, the database was already bogged down due to storing base64-encoded images, which made requests balloon to over 5MB. We solved this by switching to direct image uploads to Azure Blob Storage and returning only URLs in the response. The toughest part, though, was orchestrating multiple services and stitching everything together.
Accomplishments that we're proud of
Our biggest win? Designing an intuitive UI despite our team’s systems engineering background. Leveraging AI tools like Copilot and ChatGPT, we managed to create a seamless user experience that we’re really proud of.
What we learned
We gained a deep understanding of AI and how to integrate various Azure services, along with the challenge and excitement of designing a full-fledged product from scratch.
What's next for PeerPlates!
- Implementing a robust ratings and reviews feature to build trust between users.
- Exploring monetization options where local cottage food laws permit.
- Continuing to refine and improve the user experience with enhanced UI/UX designs.
Built With
- ai
- azure
- computervision
- fastapi
- javascript
- llms
- mongodb
- python
- react
Log in or sign up for Devpost to join the conversation.