LifeStream: AI for Thalassemia Blood Warriors

"LifeStream" is an AI-powered platform that predicts donor dropouts, re-engages them with gamified nudges, and mobilizes youth to ensure a consistent, life-saving blood supply for Thalassemia patients

Comment

Inspiration

🩸 LifeStream: AI for Thalassemia Blood Warriors

đź’ˇ Inspiration

India faces a critical blood shortage each year. National data shows that while (14.6 \, \text{million}) units of blood are needed annually, there's a consistent shortfall of nearly (1 \, \text{million}) units. For Thalassemia patients—who require lifelong, routine transfusions—this shortage can be life-threatening.

What truly inspired me was realizing that this crisis isn't only about logistics—it’s about human behavior. Donors often drop out due to a lack of motivation, engagement, or recognition. Furthermore, public understanding of Thalassemia is still low, resulting in missed opportunities for advocacy and sustained support.

The “AI for Good” hackathon and the mission of Blood Warriors inspired me to design a solution that doesn’t just connect patients and donors, but actively retains, educates, and empowers donors over time.

What it does

LifeStream is an AI-powered, gamified platform designed to create a sustainable, community-driven blood donation ecosystem. It includes three core modules:

  1. AI-Powered Donor Engagement
    • Predicts donor dropout using behavioral machine learning models trained on donation frequency, user activity, and sentiment patterns
    • Sends personalized nudges via WhatsApp/SMS and even Voice notes to re-engage potential dropouts
    • College and city dashboards promote social recognition and healthy competition
    • Allows donors to share achievements on social media platforms -Uses anonymized engagement feedback to refine prediction accuracy over time

  2. Gamification Engine

    • Rewards users with badges, streaks, and rankings to increase engagement
    • Tracks personal milestones (e.g., 1st donation, 5th donation, yearly donor)
    • Recommends ideal next donation windows based on historical patterns
    • Flags high-risk donors for proactive outreach -Includes a point system for redeemable perks like digital certificates and NGO acknowledgments

  3. Awareness & Campus Mobilization

    • Educates users with microlearning content on Thalassemia and blood donation
    • Includes a College Ambassador Program to drive youth-led campaigns
    • Provides myth-busting infographics and regionalized donor stories
    • Supports six Indian languages: Hindi, Bengali, Tamil, Telugu, Marathi, and Kannada
    • Features a multilingual chatbot to answer FAQs and promote blood literacy

How we built it

Planned Technology Stack

  • Frontend: React.js (Web), Flutter (Mobile)
  • Backend: FastAPI (Python)
  • AI/ML: PyTorch, XGBoost (deployed via Azure Machine Learning)
  • Database: Azure Cosmos DB (globally distributed NoSQL)
  • Messaging: Twilio, WhatsApp Business API
  • Hosting: Azure Functions (serverless), Azure Kubernetes Service (AKS)

LifeStream will be built as a modular, cloud-native system. Since no public donor behavior dataset exists, our initial models will be trained on simulated behavioral data. Deployment will use Azure Machine Learning and AKS for scalability. The architecture is modular to support future integrations with e-RaktKosh and Blood Bridge.

Challenges we ran into

  • Data Availability
    Lack of public datasets means we must rely on synthetic training data. Collaborations with like Blend60 may provide anonymized real-world data.

  • System Integration
    e-RaktKosh and Blood Bridge may not provide open APIs, so we need an adaptable backend.

  • Privacy & Security
    Handling personal health data demands encryption, consent flows, and secure access control.

  • Sustaining Engagement
    Designing gamification that is meaningful and ethical requires user-centered UX thinking.

Accomplishments that we're proud of

I am proud of developing a comprehensive and innovative project idea that addresses a critical social problem. LifeStream is more than an app—it’s an ecosystem designed to address donor retention, education, and engagement using AI and gamification.

I’m especially proud of the interdisciplinary nature of this solution, blending behavioral psychology with smart architecture and a deep understanding of community needs.

What we learned

This project taught me that real-world AI isn’t just about models—it’s about empathy, design, and trust. Some key lessons:

  • How to simulate behavior-based data for training ML models
  • How modular architecture helps with uncertain integrations
  • Why gamification must prioritize ethics over engagement
  • How to make AI accessible across language and tech divides

What's next for LifeStream: AI for Thalassemia Blood Warriors

Phase 1 (MVP)

Build a foundational MVP during the hackathon:

  • Dropout prediction model (simulated data)
  • Gamification leaderboard
  • Emergency request chatbot

Phase 2 (Pilot)

  • Collaborate with Blood Warriors to validate models
  • Pilot College Ambassador Program in 3–5 colleges
  • Expand educational content to 6 Indian languages

Phase 3 (Scale)

  • Full integration with e-RaktKosh and Blood Bridge
  • Collaborate with NGOs and CSR arms for community drives
  • Launch community challenges and national-level dashboards

Additional Roadmap Goals:

  • Partner with Blend60 to refine AI models with real data
  • Develop advanced analytics in the Ambassador Dashboard
  • Test social incentives via college/city-level gamified campaigns ## 🙏 Acknowledgments

Thanks to the AI for Good Hackathon for providing a platform to transform intent into innovation. LifeStream is dedicated to every Thalassemia patient, volunteer, and donor working to make India’s blood ecosystem more compassionate and consistent.

Built With

Share this project:

Updates