Inspiration
Crohn's disease is personal. Foods that work for one person may cause problems for another. We wanted to build a tool that learns from your own food and symptom history instead of relying on generic food lists.
What it does
GutGuide tracks meals and symptoms to identify personal patterns and estimate food compatibility. Users can also scan restaurant menus with Gemini to find dishes that best match their history, explore personalized recipes, and discover nearby food options.
How we built it
Built with Next.js, React, TypeScript, PostgreSQL (Tiger Data), Gemini, ElevenLabs, and Tailwind CSS. Deployed with DigitalOcean.
Challenges we ran into
The biggest challenge was making personalization useful without presenting it as medical advice. We focused on correlations, confidence levels, and transparent explanations rather than claiming that foods cause or prevent symptoms.
Accomplishments that we're proud of
We built a complete personalized food experience around 120 days of seeded meal and symptom history, including a working menu scanner that turns an ordinary restaurant menu into personalized recommendations.
What we learned
We learned how to combine time-series data, AI, and full-stack development into a product where the underlying data actually drives the experience.
Team Members
- Christopher Achkar - caa152@pitt.edu
- Harsh Kumar - harshit.k1956@gmail.com
- Jonathan Farah - jonathanfarah05@gmail.com
- Yaseen Benharrats - yaseenbenh@gmail.com
What's next for GutGuide
Better food and ingredient recognition, easier meal logging, stronger personal pattern detection, and integrations that help users better understand their own data and discuss it with their healthcare team.
Log in or sign up for Devpost to join the conversation.