Sanjeevani

Where Life Meets Technology, One Drop at a Time.

Comment

💡 Sanjeevani — Real-Time Lifelines for Thalassemia Patients

🚀 Inspiration

Every 2–3 weeks, Thalassemia patients rely on a blood transfusion — not just for health, but for survival. Yet, families often scramble at the last minute, unable to find compatible donors or reliable systems.

What if we could change that? What if we could predict donor availability, connect them in real-time, and ensure no patient waits helplessly?

This idea became personal to us — imagining our own loved ones needing such care pushed us to build Sanjeevani: a smart, secure, AI-powered platform to bridge the gap between patients, donors, doctors, and support networks.

🧠 What We Learned

  • How to design and build real-time systems with patient-centered UX
  • The power of behavioral AI models for predicting human availability
  • How to integrate with existing healthcare platforms (e-RaktKosh, NGOs)
  • Managing secure, encrypted medical communication
  • Importance of gamification and emotional storytelling for donor retention
  • Navigating ethical design for sensitive health data

🛠️ How We Built It

🧩 Tech Stack

  • Frontend: Flutter (mobile cross-platform)
  • Backend: Spring Boot (Java)
  • Database: PostgreSQL + Redis
  • AI/ML: Python (scikit-learn, XGBoost)
  • Authentication: Firebase Auth + JWT
  • APIs Integrated: e-RaktKosh, Blood Warriors' Blood Bridge
  • Deployment: Render (for MVP), AWS planned for scale

🔧 Key Features

  • Real-Time Blood Matching based on location, blood type, and urgency
  • AI-Powered Donor Prediction using donation frequency and behavior patterns
  • Gamified Donor Journey with dashboards, badges, and impact stats
  • Encrypted Doctor–Patient Communication with text, reports, reminders
  • Parent & Patient Support Hub with education and community spaces
  • Full integration with national donor databases and NGOs

⚔️ Challenges We Faced

  1. Real-World Data Scarcity

    • No access to real donor behavior datasets → we simulated donation patterns using researched intervals and habits

  2. API Complexity

    • Government APIs (e-RaktKosh) lacked developer-friendly docs → trial-and-error integration using sandbox endpoints

  3. Balancing Emotion with UX

    • Designing UI that’s both informative and comforting — especially for parents and donors — required extra testing

  4. AI Model Accuracy

    • Human behavior isn’t always predictable → we added feedback loops to improve donation likelihood predictions over time

  5. Security & Consent

    • Ensured that donors only get contacted if they opt in, and patient medical data stays encrypted & role-restricted

✅ Results & Impact (Potential)

  • 🩸 Blood requests matched in <5 minutes
  • 🔁 Donor repeat rate could improve by 50–60% using AI nudges
  • 👪 Support for over 1000+ Thalassemia families
  • 🤝 Strong collaboration with NGOs and government systems

🎯 What's Next

  • Pilot program with an NGO partner (Blood Warriors or Sankalp India)
  • Language support for Hindi, Bengali, and Tamil
  • Integration with regional blood bank APIs
  • Partnership with government health missions for scale

“Sanjeevani is more than a system. It’s a commitment to ensure no child ever waits too long for life.”

Share this project:

Updates