-
Patient, Donor and Matching Algo. (Wireframes)
-
Health Dashboard (Wireframe)
Inspiration
Last summer when I was in India, it was supposed to be a normal vacation, but what I witnessed would forever change my perspective on accessible healthcare technology. My cousins neighbor’s daughter, was diagnosed with Thalassemia major. I couldn’t really understand what that meant but learnt from my parents discussions that it should have been a manageable condition. However, this turned into a nightmare of frantic phone calls, sleepless nights, and desperate searches for blood donors every few weeks by their family and friends. I watched them spend multiple days calling relatives, posting on social media, and visiting blood banks, only to return home empty-handed. The most heartbreaking moment came when she had to be rushed to the hospital in the middle of night, because they couldn't find a donor in time for her scheduled transfusion. That night, as I saw the family's anguish, I realized that few of us are blessed to have advanced healthcare but there are many who in 21st century and in our age of smartphones and AI, are still figuring out basic, critical healthcare problems like it was 1990’s. That's when I knew I had to build something and the research data motivated me to do more.
What it does
BloodConnect.AI revolutionizes blood donation for Thalassemia patients by transforming reactive crisis management into predictive healthcare. Instead of searching for donors during emergencies, our AI-app finds patient needs 2-4 weeks in advance and matches them with compatible donors in under 15 minutes.
Main Functionality:
Predictive Matching: Machine learning algorithms analyze patient transfusion schedules, donor availability patterns, and geographic data to create optimal matches.
Emergency Response System: When urgent requests come in, AI instantly identifies the 3 most reliable donors within 5km and sends simultaneous alerts.
Integration: Seamlessly connects with Blood Warriors' Blood Bridge system and e-RaktKosh through REST APIs.
Community Awareness: Personalized educational content spreads Thalassemia awareness and recruits new donors through social sharing.
Patients, Doctors and Donor Experience: Patients get proactive scheduling and emergency backup systems Donors receive smart notifications when they're most likely to be available and see their direct impact Hospitals get predictive analytics for blood inventory planning Communities learn about Thalassemia prevention through AI-targeted awareness campaigns.
BloodConnect.AI doesn't just digitize existing processes but it reimagines blood donation as a predictive, community-driven ecosystem where technology amplifies human compassion.
How we built it
Algorithms showing actual logic and implementation and code snippets - Frontend - React.js and Backend Architecture (Java Spring Boot), AI model - pre-trained models with - Python and scikit-learn. Simple Design wireframes and Scalable Architecture.
Challenges we ran into
Balancing Simplicity with Functionality, I mitigated by starting simple. Integration - Kept facing issues but tried alternates and used Rest API to test it locally. Couldn’t get public deployment done yet. Deployment issues - ran into build with dependencies issue especially with integration and 3rd party libraries. Testing - Compile and run issues took long time than actual coding.
Accomplishments that we're proud of
Developed variant of this to begin with that’s live on GitHub - SkinAI analyzer which is a simpler basic version where I used AI pre-trained models. It does similar matching with existing database to classify images. I got the app working and live with all integrations complete. I think that was my biggest accomplishment and made me very happy given from where I started and I plan to enhance to include blood matching with advanced AI features.
What we learned
I initially thought building AI was just easy and about maximizing use of API, but this exercise taught me that architecture, testing, and fail-safes are equally important. Every prediction could impact a human life, so I learned to: Implement confidence score (predictions >90% confidence) Add human-in-loop for oversight layers for critical decisions/exceptions Build transparent algorithms that doctors can understand and trust.
What's next for BloodConnect.AI
BloodConnect.AI started as a solution to one crisis, but I now see its potential to transform healthcare delivery across the world. This hackathon isn't just about winning a competition, it's about proving that a high school girl with passion, bit of coding skills, and a lot of determination can build technology that saves lives. The future I am building- 1 Predictive healthcare that prevents crises instead of reacting to them 2 AI-powered compassion that connects communities in their time of need 3 Technology with purpose that proves innovation can have a heart
Every algorithm I write, every feature I build, and every user I onboard brings us one step closer to a world where no patient has to wait in fear, wondering if help will come in time.
Today, BloodConnect.AI is just an app. Tomorrow, it will be the foundation of how we approach “predictive and accessible” healthcare. As they say in India, it takes village to raise the baby… I am hopeful my idea will help save lives.
Log in or sign up for Devpost to join the conversation.