BloodLink AI

Inspiration

India faces a chronic shortage of blood and a rising number of Thalassemia patients requiring regular transfusions. Inspired by real stories of patients struggling to find timely donors, we envisioned BloodLink AI — a platform to bridge the gap between blood donors, recipients, and healthcare support using the power of Artificial Intelligence.

Our goal is aligned with the mission to eliminate Thalassemia in India by 2035 through awareness, early detection, and a smarter donation network.

What it does

BloodLink AI is an AI-powered platform that:

1. Matches blood donors with patients based on location, blood type, and urgency using a smart recommendation engine. 2 . Predicts future blood demand in specific regions using time-series forecasting. 3 . Offers early risk detection for Thalassemia using symptom-based input and ML classification. Provides a dashboard for NGOs/hospitals to manage donors and raise awareness.

How we built it

Frontend: Built using Streamlit for quick UI deployment. Backend: Python with Flask APIs for handling data logic and ML integration. Machine Learning Models:

1. Classification for donor-patient matching. 2.Time-series model (ARIMA) for blood demand prediction. 3.Logistic regression for Thalassemia risk estimation.

Database: SQLite for storing donor and request data. Hosting: Deployed on Render and GitHub Pages for demo access. Tools/Libraries: pandas, scikit-learn, matplotlib, seaborn, streamlit, flask

Challenges we ran into

Data scarcity: Limited access to real Thalassemia and blood donation datasets. Balancing complexity vs speed in ML model training and response time. Building a unified UX for hospitals, donors, and patients within one app. Ensuring accuracy in health-related predictions while avoiding misinformation.

Accomplishments that we're proud of

Developed a working prototype with donor matchmaking and demand prediction. Built a clean and interactive dashboard using Streamlit. Managed to simulate real-life blood bank data and train usable models on it. Addressed a real social issue with tech that can scale

What we learned

1. Applied AI/ML for social good in the healthcare space.
2. Importance of clean data preprocessing for health-focused models.
3. Improved our collaboration and project management using GitHub. 4.Gained confidence in building end-to-end AI solutions under time constraints.

What's next for BloodLink AI

Partnering with blood banks, NGOs, and hospitals for pilot testing. Adding a mobile-first version to reach remote users. Integrating real-time location tracking and donor notification system. Expanding ML models to include risk prediction for other blood disorders. Launching a donor awareness campaign with gamified incentives.

Built With

• flask
• matplotlib
• pandas
• render
• scikit-learn
• seaborn
• sqlite
• streamlit
• ython