ThalAI

ThalAI – Empowering Lives with Every Drop

Comment

ThalAI – Empowering Lives with Every Drop

Inspiration

While volunteering at a Thalassemia awareness camp, we met families who had traveled over 100 km just to receive a scheduled blood transfusion — only to return home because no blood was available. This experience made us realize that the problem isn't just about supply; it's about timing, prediction, and communication.

Thalassemia patients need recurring transfusions. But India lacks a proactive system to predict their needs and prepare donors ahead of time. This gap inspired us to create ThalAI — an intelligent, preventive care platform that uses AI, voice technology, and low-cost IoT to support patients and their families.

What it does

ThalAI is a unified platform that provides:

  • Transfusion Risk Prediction
    Predicts the next transfusion date using AI models trained on patient history, symptoms, and vitals.

  • IoT-Enabled Vitals Monitoring
    A low-cost wearable or mobile sensor reads pulse and temperature. When vitals drop below normal, the system issues early warnings to family members and local hospitals.

  • Voice-Based Virtual Assistant
    Enables caregivers to ask health-related questions in Hindi, Gujarati, or Marathi using voice. Answers are personalized and verified by medical experts.

  • Blockchain-Backed Medical Timeline
    Stores transfusions, lab reports, and consultations securely. Patients control who accesses their data.

  • Proactive Donor Engagement
    Predicts availability of nearby donors based on past donation patterns and notifies them before a crisis.

How we built it

  • Frontend: React Native and Flutter for the mobile app, with a WhatsApp chatbot fallback for low-tech environments.
  • Backend: FastAPI and Firebase for secure storage and real-time communication.
  • AI Models:
    • LSTM and XGBoost for transfusion risk prediction.
    • BERT and fine-tuned speech-to-text for regional language voice assistant.
  • IoT Prototype: NodeMCU with a pulse sensor and thermometer, optionally connected via mobile phone or standalone.
  • Blockchain: Used Hyperledger/IPFS to simulate an immutable patient health ledger with consent-based access control.

We used synthetic patient data based on real-world patterns and WHO data to train and test our models.

Challenges we ran into

  • Lack of publicly available Thalassemia patient data for model training.
  • Building an inclusive voicebot that understands regional dialects with low noise tolerance.
  • Designing an interface for families with low digital literacy.
  • Ensuring affordability and offline functionality in the IoT module.
  • Balancing medical data privacy with seamless communication between patients, hospitals, and donors.

Accomplishments that we're proud of

  • Developed a unique Transfusion Risk Scoring Engine based on real-world patient patterns.
  • Built a multilingual voice assistant that can operate offline and help low-literacy users.
  • Created a secure medical data ecosystem where patients fully control their information.
  • Designed a proactive system that alerts donors before a critical request is raised.
  • Integrated AI, voice technology, and IoT into a working, field-adaptable solution.

What we learned

  • Preventive care is far more effective than reactive systems, especially in chronic conditions like Thalassemia.
  • For many families, literacy is not just about reading — it's about understanding and trusting the healthcare system.
  • Voice and audio-first interfaces can bridge the accessibility gap where traditional apps fail.
  • AI needs to be explainable, especially in medical contexts where outcomes affect lives.
  • Strong data governance is essential in any healthcare solution involving patients and real-time decisions.

What's next for ThalAI

  • Partner with Blood Warriors and government hospitals to conduct pilot testing.
  • Add image-based support for prescription recognition and medicine identification.
  • Expand language support to include Tamil, Bengali, and Kannada.
  • Deploy the IoT monitoring module to rural health centers and homes with CSR-backed funding.
  • Collaborate with ABHA (Ayushman Bharat Health Account) for integrated patient health ID linkage.
  • Open up APIs for NGOs, hospitals, and community health workers to integrate with ThalAI services.

Built With

Share this project:

Updates