Pulse Match

Inspiration:

We were moved by the heartbreaking reality that over 10,000–15,000 Indian children born with thalassemia major each year depend on life-saving transfusions—and often face severe shortages. Seeing organizations like Blood Warriors tirelessly drive awareness, we still saw a tech gap: why can’t AI match donors to patients as fast as ride-hailing apps match riders? That “aha” moment sparked Pulse Match —an Uber-style, AI-powered platform to sync donors and thalassemia families, plus give them ongoing support and education.

What it does:

• Real-time Donor Matching: Patients (or caregivers) request blood by type/urgency; our AI engine instantly finds nearby, eligible donors and sends them push alerts.

• Predictive Availability: A machine-learning model scores donors on readiness (days since last donation, health flags, past frequency) so we notify the likeliest givers first.

• Patient Support Portal & Chatbot: Secure messaging/video with specialists, automated transfusion & test reminders, and an AI chatbot answering care FAQs (nutrition, chelation schedules).

• Gamified Donor Dashboard: Donors earn “Life Saver” badges, track total donations, and view impact stats—turning once-off givers into recurring heroes.

• Admin Analytics: Hospitals and NGOs get live stats on requests, donor activity, and forecasted blood demand to optimize outreach.

How we built it:

Frontend: React.js + Material UI for a slick, Amazon-style UX.
Backend: Node.js + Express serving REST APIs.
Database: PostgreSQL to store users, donations, patient records.
Machine Learning: Python (scikit-learn) prototype predicting donor readiness from synthetic data.
Chatbot: Azure Bot Service with curated medical FAQs (rule-based MVP).
Mapping & Alerts: Google Maps API for geolocation; Twilio for SMS/push notifications.
Security: OAuth 2.0, JWT, TLS encryption, compliance with India’s DPDP Act for health data.

Challenges we ran into:

• Data Scarcity for ML: Real donor histories are sensitive. We used anonymized samples and synthetic data to train our MVP model, planning to refine with actual logs later.

• Healthcare Compliance: Balancing HIPAA-grade security with India’s privacy laws meant extra work on consent flows, encryption, and legal disclaimers.

• API Integration Roadblocks: e-RaktKosh’s government APIs required mocks in the hackathon, but we mapped out clear steps for full integration post-event.

• UX for Low-Tech Users: Rural families and patients have varied digital literacy—so we prioritized one-click flows, multi-language support, and chatbot assistance.

• Medical Liability: Ensuring chatbot advice stays informational, not diagnostic, with visible disclaimers and expert-verified content.

Accomplishments that we're proud of:

• 🚀 48-Hour MVP Demo: Live donor-patient match workflow, AI scoring prototype, and basic chatbot all in one polished web app.

• 🎨 Enterprise-Grade UI: A responsive, intuitive interface inspired by Amazon/Microsoft, even for our hackathon build.

• 🤖 Proof-of-Concept AI: Early predictive model showing promise in donor readiness accuracy.

• 💬 Chatbot Integration: A working rule-based bot answering key thalassemia care questions in real time.

• 🤝 Scalable Architecture: Modular design ready to add new features (genetic screening, mobile PWA) and expand beyond thalassemia.

What we learned:

• A single unit of donated blood can save up to three lives—but only if matched quickly.

• Combining logistics (donor matching) and care (telehealth + education) delivers far more impact than isolated tools.

• Real-world health projects need airtight data-privacy and compliance strategies from day one.

• Even a minimal AI prototype drives home the value of predictive alerts in volunteer engagement.

• Simplicity in design is crucial for diverse user groups, especially those with limited tech access.

What's next for Pulse Match:

Data & Model Refinement: Integrate real donation logs to retrain and improve our predictive engine.
e-RaktKosh & Partner APIs: Full integration with national blood-bank data and Blood Warriors’ donor registry.
Enhanced Chatbot AI: Layer in LLM-powered NLP for richer, more conversational patient support.
Mobile PWA & Offline Mode: Reach rural areas with low connectivity via an offline-first app experience.
Pilot Programs: Launch in 1–2 hospitals/blood banks to gather live feedback and demonstrate impact.
Scale Nationwide: Add regional languages, onboard more NGOs/hospitals, and evolve into a universal “Uber for Blood.”

Together, we’ll keep syncing hearts—one drop at a time. ❤️