Inspiration
A couple of years ago, my brother had a serious accident. Our family was desperate to get blood for him, but the process was chaotic and slow. We called hospitals, searched lists, and reached out to everyone we knew—still, it felt impossible to get the right blood match quickly. That helpless feeling, watching time slip by in a life-or-death situation, stayed with me. I decided there had to be a better way for families to get lifesaving blood quickly and reliably in emergencies. That experience is what inspired me to build this project, aiming to help others avoid the same fear and stress my family went through.
What it does
Our project is a secure, AI-powered chatbot platform for blood donation and Thalassemia care. It streamlines emergency blood requests, matches donors and recipients instantly, and provides real-time education and self-management tools for Thalassemia patients. The platform intelligently engages donors, manages communication for all stakeholders, and keeps everything safe with robust security and privacy measures—all from an easy-to-use web dashboard.
How we built it
We started by mapping out the real-world problems and user stories from both personal experience and healthcare research. The frontend is built with React, providing a fast, friendly interface for patients, donors, and admins. For conversational AI, we integrated large language models (LLMs) using Groq’s super-fast inference API, so all education, FAQs, and symptom checking are handled naturally in chat. Our backend uses Python (FastAPI/Flask) for the main API and business logic, with Node.js powering real-time WebSocket notifications for emergencies. MySQL stores structured data (users, patients, donors), and MongoDB handles all chats and flexible knowledge base content. We implemented OAuth2/JWT for authentication, ensured TLS/SSL encryption, and included 2FA for sensitive accounts. Everything is containerized with Docker, deployed on the cloud, and monitored using industry-standard DevOps tools.
Challenges we ran into
- Making sure urgent blood requests get matched and sent instantly to the right donors was technically tricky. We had to ensure the messaging worked in real time with no leaks or delays.
- Balancing strict security and privacy (HIPAA, GDPR compliance) while keeping the user experience smooth was a constant challenge.
- Translating detailed clinical guidelines into clear and safe chatbot conversations, and then testing them for accuracy, took a lot of time and collaboration.
- Getting LLM responses through Groq to feel both instant and reliable—with useful, non-generic answers—required careful backend design and plenty of trial and error.
- We had to simulate lots of real-life scenarios: emergency surges, consent changes, partial downtimes, and data failures, to make sure everything works when it’s really needed.
Accomplishments that we're proud of
- Seeing the system smoothly match emergency blood requests to real donors in test environments—cutting through chaos and anxiety.
- Creating a chat experience that actually educates and reassures Thalassemia patients, based on best clinical practices.
- Building an easy-to-use, unified interface that helps donors, patients, and hospitals communicate and take action instantly.
- Achieving strong security standards without making the system hard to use.
- Turning a deeply personal experience into something that could have a positive, real-world impact for others in crisis.
What we learned
- The best solutions come from understanding real, lived pain points—not just tech trends.
- Automating and securing healthcare processes is tough but vital. Communication gaps can be life-threatening, so clarity and reliability matter.
- Strong privacy and compliance principles don’t have to make systems rigid; with planning, they can be built in from the start.
- It’s possible to make AI chat meaningful and trustworthy (not just flashy), if built thoughtfully and validated with actual medical input.
- Working on this project made us more empathetic to what families and frontline health workers actually go through during emergencies.
What's next for Untitled
Our next steps are:
- Onboarding real hospitals and blood banks as pilot partners, to test the platform in live settings.
- Expanding our chatbot’s medical coverage—for more chronic conditions and language support.
- Integrating government and NGO donor databases to widen reach and reduce mismatches.
- Adding more advanced analytics and “smart” automation to anticipate needs before they become urgent.
- Building a mobile-friendly version to reach even more users.
- Gathering real feedback from patients, donors, and staff to keep improving usability, safety, and reliability.
We’re excited to keep pushing the project forward, with the aim of making emergency blood access fast, trustworthy, and stress-free for everyone.
Log in or sign up for Devpost to join the conversation.