Tracky.health - Stop Googling Your Symptoms

$4.5T healthcare industry runs on guesswork. We're fixing that. Tracky.health uses AI to connect every health datapoint and predict what doctors miss. Your personal health oracle.

Comment

Inspiration

My wife is health-conscious (okay, maybe a bit hypochondriac 😅), and every trip we take involves packing a stack of medical documents thicker than a phone book. Lab results from different clinics, prescription notes in illegible handwriting, test results spanning years – all because "what if we need them?"

Watching her Google symptoms and try to connect dots between scattered health data made me realize: we're living in 2025 but managing health like it's 1995. The $4.5T healthcare industry runs on guesswork because critical health information is trapped in silos, and patients are left to be their own advocates with incomplete, incomprehensible data.

What it does

Tracky.health is your personal health oracle that connects every health datapoint to predict what doctors miss:

  • AI Document Parser: Upload any medical document (lab results, diagnoses, prescriptions) and AI extracts structured data instantly
  • Health App Integration: Connect Strava to sync activity data
  • Smart Lab Tracking: View parameter history with trends and export reports as PDF
  • AI Health Chat: Ask questions about all your parsed health data with context-aware responses
  • Daily AI Insights: Get personalized health recommendations and predictions every 24 hours

Instead of Googling "why am I tired," users can ask "My energy dropped 23% this month – what's the connection with my recent lab results?"

How we built it

Built from scratch over the course of a month, working evenings and weekends around my 9-to-5 job, primarily using Bolt.new as my development platform:

  • Frontend: Next.js with medical-grade UI design
  • Backend: Supabase for real-time data and authentication
  • AI Integration: Google Gemini for document parsing and health insights (planning to experiment with other models)
  • Document Processing: Google Gemini
  • Payments: Stripe

Bolt.new was a game-changer for rapid prototyping and iteration – allowing me to build complex features while juggling full-time work.

Challenges we ran into

  • Deploy Challange - Application deployed to Netlify
  • Startup Challenge - Application uses Supabase for database management, user authentication, and storing documents

Accomplishments that we're proud of

  • Actually finishing it: I had a lot of doubts and fears throughout the process, but I pushed through and completed a fully functional health platform
  • Real user validation: My wife tried the service as an actual user and she genuinely likes it – the best validation I could ask for
  • Mastering prompt engineering: It was fascinating to work with AI prompts to get better results from document parsing to health insights generation
  • Learning Bolt.new: I'd never used Bolt.new before this project, but I figured it out and it became an essential tool for rapid development
  • User Empowerment: Transforming "I feel tired sometimes" into "My energy levels declined 23% over the past month, coinciding with iron levels dropping below optimal range"

What we learned

The healthcare system is fundamentally broken by data silos. Patients generate massive amounts of health data – smartwatches, lab tests, medical records – but it's all disconnected. Doctors make decisions with incomplete information because relevant data is trapped in different systems.

AI isn't just a buzzword here – it's a necessity. The human brain can't process and correlate hundreds of health datapoints across months or years, but AI can spot patterns that even experienced doctors miss. Most importantly: users don't want another health app. They want health intelligence that actually improves their lives and medical conversations.

What's next for Tracky.health

  • Expand Integrations: Apple Health, Google Fit, sleep trackers, and major EHR systems
  • Advanced Predictions: Machine learning models for early disease detection and health risk assessment
  • Provider Integration: Tools for sharing insights with healthcare providers for better consultations
  • Community Features: Anonymous health pattern sharing to improve AI predictions
  • Mobile App: Native iOS/Android apps for document scanning and real-time health monitoring

We're not trying to replace doctors – we're giving patients the tools to have dramatically better conversations with their healthcare providers. This is the future of personalized healthcare, where your health data works for you, not against you.

From paper chaos to health clarity – one data point at a time. 🚀

Built With

Share this project:

Updates