🩺 HealthNarrator – Project Story

💡 Inspiration

We’ve all seen it — a loved one receives a medical report, their eyes glaze over the numbers, and they’re left with more anxiety than answers. Whether it's our grandparents in rural areas or friends who skip follow-ups out of confusion, we realized there’s a huge gap in health understanding. We were inspired to build HealthNarrator to bridge that gap — turning complex medical reports into human, compassionate, and actionable health stories anyone can understand.

🤖 What It Does

HealthNarrator is an AI-powered platform that:

  • Accepts PDF or image-based medical reports.
  • Explains each medical term and value in plain, friendly language.
  • Highlights potential health issues and when action is needed.
  • Gives detailed lifestyle guidance — diet, sleep, hydration, follow-ups.
  • Supports translation into local languages, making it accessible to all.

It’s like having a personal health interpreter, available anytime.

🛠️ How We Built It

We used:

  • Django for the backend and file handling.
  • OpenAI’s GPT-4 API to process and explain medical text.
  • OCR (pytesseract) and PyPDF2 to extract text from images and PDFs.
  • Clean, HTML/CSS frontend styled with a warm, approachable theme.
  • Markdown-to-HTML rendering for rich, readable output.
  • Language translation through prompt chaining with GPT.

Each feature is focused on clarity, empathy, and empowerment.

🚧 Challenges We Ran Into

  • Parsing medical data accurately from low-quality images was tough.
  • GPT prompts needed careful tuning to maintain accuracy without overwhelming the user.
  • Displaying markdown responses with proper styling took refinement.
  • Balancing detail with simplicity — too much info can still confuse users.
  • Local language translation had to be handled delicately to maintain context and tone.

🏆 Accomplishments That We're Proud Of

  • Built a fully functional prototype that works seamlessly from upload to explanation.
  • Created a truly intuitive UI that anyone — even non-tech-savvy users — can navigate.
  • Successfully implemented dynamic, user-defined translation into any major language.
  • Got positive feedback from test users who felt more confident about their health.

📚 What We Learned

  • Simplicity is powerful. The less users have to think, the more they understand.
  • AI can be an incredible health ally when guided correctly.
  • Building inclusive solutions (language, design, clarity) creates 10x more impact.
  • The real challenge isn’t just tech — it’s trust and empathy in the experience.

🚀 What's Next for HealthNarrator

  • Launch a mobile app for greater accessibility in rural and underserved areas.
  • Add voice-based explanations for the visually impaired.
  • Improve Security by ensuring that user data is never conpromized
  • Partner with clinics and NGOs to bring this tool to the people who need it most.
  • Expand translation to support regional dialects and voice-activated assistants.

Our vision: To make every health report not just readable — but empowering.

Built With

Share this project:

Updates