RealtimeBloodCall

🧠 Inspiration

Thalassemia patients in India often face life-threatening delays in getting the right blood at the right time. Existing solutions are fragmented, language-limited, and reactive. We wanted to create something that thinks ahead—an AI-powered system that doesn’t wait for a crisis but acts before one.

💡 What it does

RealtimeBloodCall uses a network of AI agents to connect Thalassemia patients with suitable donors in real-time, across languages and locations, while integrating prediction, outreach, and verification into one seamless system. Key features:

• Predicts donor availability based on behavior and proximity
• Sends multilingual alerts automatically via SMS/WhatsApp/app
• Coordinates outreach autonomously using AI agents
• Tracks donor eligibility, reliability, and past donations
• Logs verified donations using blockchain smart contracts
• Supports patients and families in 13+ Indian languages

🏗️ How we built it

• AI Agents (Azure Agent Orchestrator+ GPT-4) orchestrate donor search, prediction, outreach, and multilingual support
• React + Next.js + Tailwind for a fast, responsive frontend
• RainbowKit + Base (Sepolia) for secure, verifiable donor IDs and reward tracking (Will be tailored to the learning curve of the user)
• Azure Functions powers serverless logic with agent integrations
• IPFS + Redis + Smart Contracts store donation logs and confirm actions
• X402 Protocol supports token-based reward & payment incentives
• Language AI bridges urgent messages across Hindi, Kannada, Tamil, Telugu, Bengali, and more

⚙️ Challenges we ran into

• Getting AI agents to coordinate without overlap or delay
• Handling nuanced regional language translation for medical terms
• Avoiding donor fatigue with smart notification throttling
• Making blockchain efficient and unobtrusive for health users
• Ensuring low-latency AI inference across mobile networks

🏆 Accomplishments we’re proud of

Taking up a problem this real, this urgent, and this overlooked—that in itself is our biggest achievement. We didn’t just brainstorm a hackathon idea. We committed to building something that could genuinely save lives for Thalassemia patients across India.

We’re proud to have shaped a solution that blends AI agent orchestration, regional language inclusivity, and ethical healthtech design from day one. And above all, we’re proud to say: we’re just getting started—this will become the best platform of its kind, built with purpose and people at its core.

📚 What we learned

• Real-world AI needs context, culture, and clarity
• Tech isn’t enough—timing, language, and trust matter most
• Agent orchestration is powerful when roles are well-defined
• Web3 + health needs to be lightweight and user-first
• Real impact comes when AI thinks like a human, not just a machine

🚀 What’s next for RealtimeBloodCall

• Agent-Driven Architecture: We’ll deploy multiple Azure-powered AI agents with clearly scoped roles—prediction, outreach, verification, and language mediation—running via Azure functions with agent routing logic.

• Multilingual First Approach: We'll fine-tune models and use Azure Translate + GPT Titan embeddings for seamless regional language support in over 13 Indian languages, including personalized patient-donor messages.

• Realtime Donor Matching Engine: Built using Redis and GeoFire to match location, blood type, and eligibility within seconds—powered by AI inference and past donation data.

• Secure & Transparent Backend: Using Base (L2) and IPFS to log donations securely, paired with RainbowKit wallets for donor identity, reputation, and future reward systems.

• Mobile-Optimized Web App: Built with Next.js and Tailwind for responsive UI, with real-time notifications through WhatsApp Cloud API and SMS.

• Pilot Testing: Launching initial pilots with Blood Warriors and local hospitals in Andhra Pradesh and Telangana, followed by scaled rollout via integrations with e-RaktKosh and Indian Red Cross.

• National rollout with e-RaktKosh, hospitals, and NGOs

• Launch of Donor Karma NFTs to reward verified donations

• Predictive dashboards for hospitals and blood banks

• Tighter WhatsApp + UPI integration for accessibility and donations

• A public API layer for third-party health apps to plug into our AI agents

Built With

• azure-api-management
• azure-blockchain-service
• azure-cache-for-redis
• azure-cognitive-services-translator
• azure-functions
• azure-notification-hubs
• azure-openai-service-&-semantic-kernel
• framer-motion
• ipfs
• next.js
• tailwind-css
• whatsapp-cloud-api