Guthealth.chat

Inspiration

Guthealth.chat was born from personal curiosity and frustration with the overwhelming and sometimes contradictory advice on gut health found online. We wanted an approachable, conversational AI buddy that could deliver reliable, science-backed answers while making gut health less intimidating and more actionable for anyone.

What it does

Guthealth.chat is an AI powered web app that answers your gut health and nutrition questions instantly. It pulls from a curated knowledge base, interprets your queries using powerful language models, and serves up clear, insightful advice—no jargon, no judgement. Whether you’re wondering about probiotics, symptoms, diets, or just want practical tips, it’s your smart companion for digestive wellness.

How we built it

We combined two main tech stacks:

Backend: Python, FastAPI, and libraries like transformers, langchain, and faiss-cpu. We scrape and clean gut health data, store it in structured formats, and create embedding-based search so responses are relevant and factual.
Frontend: React, with a clean UI for chatting. The frontend sends your queries to the backend, which runs LLM inference and returns responses in real-time.

Challenges we ran into

Getting reliable, scientifically vetted data on gut health was harder than expected.
Making the AI respond conversationally not too robotic, but not overly casual.
Ensuring speed while keeping accuracy high: optimizing the backend and search pipeline was technically tricky.
Integrating frontend and backend so the chat felt seamless.

Accomplishments that we're proud of

Built a conversational AI that really “gets” the user offering not just generic advice, but tailored, actionable info.
The integrated data pipeline: from raw scraping to cleaned, usable information.
A simple user experience no signups, no ads, just answers.

What we learned

Gut health is deeply personal, and most people just want clear, digestible facts.
Balancing simplicity and reliability in AI responses is key for trust.
Building the bridge between technical AI workings and real human needs was challenging, but super rewarding.