📖 About the Project 💡 Inspiration In real-world AI and healthcare innovation, access to quality, privacy-preserving datasets is a huge barrier. Strict regulations like HIPAA and GDPR make it difficult for researchers and startups to build robust AI models with real patient data.
That's where HealthSynthAI comes in.
We wanted to solve this by creating a lightweight, ethical, and customizable synthetic data generator specifically for the healthcare domain — so AI research and model training can thrive without compromising patient privacy.
🛠️ How We Built It We developed HealthSynthAI using Streamlit to build a clean and interactive frontend. Users can:
Input key features like age, gender, medical conditions, and record count
Instantly generate realistic synthetic health records
Preview and download results in CSV format
The logic was built in Python, and data was structured using Pandas. We added Lottie animations to humanize the UI and included styling with custom CSS for a modern look.
🚧 Challenges We Faced Deploying Python-based apps on free platforms (avoiding AWS/credits)
Ensuring the synthetic records were believable yet privacy-safe
Optimizing for fast UI updates with no backend dependencies
Keeping the design polished, interactive, and hackathon-ready
🎯 What We Learned Building humanized GenAI apps without needing cloud credits
UI/UX design with minimal tools using Streamlit and Lottie
Deploying apps on Streamlit Cloud without backend config
Creating realistic synthetic data structures with ethical constraints
Log in or sign up for Devpost to join the conversation.