HemeLink

HemeLink's AI ends the scramble for blood. It proactively finds reliable donors for Thalassemia patients, ensuring timely transfusions and peace of mind.

Comment

HemeLink - A Smarter Way to Connect Patients and Donors

Team Name: HemeLink

Institution: VIT, Chennai

Project Members: Raghul, Shyam Santhosh

Hackathon: AI for Good Hackathon 2025

Inspiration

In India, thousands of Thalassemia patients depend on timely, compatible blood donations - often every 2-4 weeks. Despite the existence of platforms like e-RaktKosh and Blood Bridge, the process of finding available donors in real time remains frustrating and impersonal.

We imagined a more empathetic, intelligent, and connected system - one where:

  • Patients could talk to a chatbot in their language,
  • Donors could be predicted even before they registered,
  • Doctors could monitor communications, and
  • Both patients and donors feel emotionally involved.

HemeLink was born from this vision - a secure, AI-driven, multilingual, gamified platform connecting lives with purpose.

What it does

HemeLink offers a comprehensive ecosystem to solve the donor-patient connection gap with intelligence, emotion, and accessibility at its core.

Core Features

Real-time Chatbot for Patient-Donor Matching
A multilingual AI chatbot helps patients or caretakers find matching donors instantly, using live data and predictive analytics. The entire interaction happens in their preferred language via text or voice.

Predictive Donor Availability Insights
We use a machine learning model that factors:

  • Past donation records
  • Age, gender, and location
  • Frequency and eligibility windows
    to predict donor availability. This helps patients plan and reduces emergency panic.

Multilingual Voice/Text Communication
Patients and doctors can talk using voice/text in their own language, with instant translation. This removes barriers in communication and empowers regional or rural patients.

Gamified Donor Dashboard
A motivational, leaderboard-style dashboard showing:

  • Top Donors in India
  • Local Hero of the Month
  • Lives Saved
  • Animated human-body visual showing amount of blood donated
    Donors can track their impact, earn virtual badges, set donation reminders, and receive secure, personalized thank-you messages from patients, increasing emotional involvement and recurring participation.

Thalassemia Education Module
This chatbot allows access to animated guides, lifestyle tips, and FAQs in regional languages to help educate patients and caregivers in a more accessible and interactive way.

Consent-Based Secure Data Storage
Two possible implementation modes:

  • PostgreSQL with encrypted fields, access tokens, and role-based authorization
  • Blockchain using Ethereum smart contracts and IPFS to securely store patient-donor consents and actions

Integration-Ready APIs
We’ve built modular APIs that can plug into existing government systems like e-RaktKosh and Blood Bridge, ensuring minimal friction once integration is approved by the organizing body.

How we built it

  • Frontend: React.js + TailwindCSS (Responsive, Accessible)
  • Backend: FastAPI
  • Database: PostgreSQL (optionally backed by Ethereum Smart Contracts + IPFS)
  • AI/ML: Donor prediction model using Python (Scikit-learn + Pandas)
  • Translation: Google Cloud Translation API for NLP-based multilingual support
  • Real-time: Socket.io and WebSockets for live chatbot and dashboard updates

Challenges we ran into

  • Building a trustworthy system where people feel safe sharing sensitive medical and donor data
  • Ensuring accuracy in donor prediction without access to real blood bank APIs
  • Creating emotionally resonant experiences like visual feedback, secure messaging, and gamified engagement
  • Handling multilingual text-to-speech and speech-to-text effectively for underrepresented Indian languages

Accomplishments that we're proud of

  • Designed a fully functional donor prediction engine
  • Created a gamified experience with emotional incentives
  • Prototyped a voice+text multilingual chatbot
  • Drafted secure, integration-ready APIs based on national healthcare system formats
  • Maintained a human-first design throughout our UX journey

What we learned

  • Empathy is non-negotiable when designing healthcare solutions
  • Privacy, transparency, and data ownership are core to patient trust
  • Gamification and communication features are powerful drivers of donor retention
  • Accessibility (regional languages, visuals, simplified UI) is a key differentiator

What's next for HemeLink

Phase Timeline Goals
During Hackathon 2 Days Build MVP with chatbot, dashboard, basic ML prediction
Post-Hackathon 1 Month Pilot with patient/donor groups, gather real-world feedback
Short Term 2–3 Months Integrate with national platforms (once approved by organizers)
Long Term 6+ Months Launch full-scale mobile app, expand to other use cases (e.g., SCD)

Built With

Share this project:

Updates

Submission history