Project BloodLink AI

Our Inspiration: A Lifeline for Warriors Our journey began with a stark reality faced by thousands of Thalassemia patients across India. These "Blood Warriors" and their families are locked in a relentless cycle: a critical need for blood transfusions every few weeks, followed by a frantic, high-stress search for donors. We were inspired by the tireless work of non-profit organizations like Blood Warriors, who operate on the front lines of this crisis. We saw that their heroic efforts were often hampered by a reactive, fragmented system. The problem wasn't a lack of willing donors; it was the chasm between a patient's need and finding the right donor at the right time. We knew that technology, specifically AI, could be the bridge across that chasm. Our inspiration was to build not just an app, but a true lifeline—a system that brings predictability and peace of mind to these families.

What We Learned: From a Simple Idea to a Smart System This project has been a profound learning experience. We quickly learned that the challenge was far more complex than just creating a donor database.

The Power of Proactive vs. Reactive: Our initial thought was to simply build a "search engine" for blood. We learned that the real innovation lies in being proactive. This led to the core idea of our Predictive Donor Availability Engine, which forecasts potential shortages, allowing the system to act before an emergency arises.

User-Centric Design is Everything: We learned that for a platform like this, trust and ease-of-use are paramount. This insight guided our decision to create a two-step donor onboarding process. By minimizing the initial friction, we maximize sign-ups, and by making location details a mandatory second step, we ensure the quality of our donor network.

AI as an Enabler of Empathy: We learned that AI's true power isn't just in raw computation, but in its ability to enable a more intelligent and empathetic system. Our Donor Suitability Score (DSS) is a testament to this. Instead of blasting every donor with notifications, the AI helps us find the most suitable person, respecting the donor's time and preventing "notification fatigue." The formula we developed is:

DSS=w 1⋅P avail+w 2 ⋅D prox+w 3⋅Hrec ​ Where we weigh predicted availability (P avail ), proximity (Dprox ), and historical reliability (H rec) to make the best match.

How We Built It: A Sprint of Collaborative Development Given the urgency of the problem, we adopted a rapid, agile development approach. Our process was a focused sprint:

Foundation First: We began by establishing a solid technical foundation, setting up a git repository on GitHub for version control and defining our project's core folder structure (backend, frontend, docs).

Defining the MVP: We aggressively scoped our Minimum Viable Product (MVP), focusing on three core user stories: Donor Onboarding, Patient Blood Request, and Donation Logging. This prevented feature creep and kept us focused on delivering immediate value. Ideas like GPS navigation were placed in a "Phase 2" roadmap.

User-Driven Architecture: We translated our user stories into concrete wireframes and technical requirements. This user-first approach dictated our database schema (Users, BloodRequests, Donations) and the API endpoints we would need.

Choosing the Right Tools: We selected a modern, scalable tech stack: Flutter for a cross-platform mobile app, Python for the AI-driven backend, Node.js for real-time communication, and PostgreSQL for our database.

Challenges We Faced: Overcoming Hurdles as a Team No project is without its challenges, and ours was no exception.

Initial Technical Hurdles: Even setting up the basics, like configuring a local Git environment, presented initial challenges. We worked through them collaboratively, which strengthened our communication and problem-solving process.

The Scoping Dilemma: One of our biggest challenges was deciding what not to build for the MVP. The temptation to add more features was strong, but we remained disciplined, focusing on the core user journey to ensure we could deliver a functional and reliable product quickly.

Balancing Simplicity and Data: We debated how much information to ask from donors upon sign-up. The challenge was to gather enough data to be useful (like location) without overwhelming the user and causing them to abandon the process. Our two-step solution was our answer to this critical design challenge.

Through this journey, we have not only architected a powerful solution but also built a robust framework for tackling complex, real-world problems with focus, collaboration, and a deep sense of purpose.

Built With

• ai-/-machine-learning:-tensorflow
• apis-&-integrations:-google-maps-api
• backend:-python-(with-flask/django)
• cloud-&-devops:-google-cloud-platform-(gcp)-/-amazon-web-services-(aws)
• dart-*-**backend:**-python-(with-flask/django)
• database:
• docker
• e-raktkosh-api
• frontend-(mobile-app):-flutter
• gemini
• github-*-**apis-&-integrations:**-google-maps-api
• node.js-(with-express.js)-*-**ai-/-machine-learning:**-tensorflow
• pandas
• pandas-*-**database:**-postgresql-(with-postgis-for-geospatial-queries)-*-**cloud-&-devops:**-google-cloud-platform-(gcp)-/-amazon-web-services-(aws)
• postgresql
• pytorch
• scikit-learn
• twilio