🌟 Inspiration The inspiration for MedSplit came from a series of personal and eye-opening experiences.
A few months without health insurance left one of our team members unable to refill insulin. Their A1C jumped dangerously from 8 to 12, showing just how quickly costs can threaten lives.
Another time, a two-month supply of medicine went completely to waste when a prescription was changed. The nearest donation drop-off was too far, with a process too complicated to be practical.
We also saw how bulk buying lowers per-unit costs, yet patients rarely have the financial capacity to commit to large upfront purchases on their own.
These experiences made one thing clear: the system leaves people struggling to access essential care, while valuable medicine is wasted. That reality inspired us to create MedSplit — a platform to bring communities together, lower costs, and ensure medicine reaches those who need it most.
📚 What We Learned
Building MedSplit taught us lessons beyond just coding:
- Healthcare empathy: Access to medicine isn’t only about affordability — it’s about equity, trust, and dignity.
- Design for scale: Even a small feature, like a demand bar for bulk orders, requires careful thought to scale across communities.
- APIs in practice: We explored how to integrate Commure for healthcare innovation, Capital One for financial inclusion, and Gemini API for breaking language barriers.
- The power of community: Simple acts — like donating unused medicines or contributing as little as $20 as a micro-grant — can change someone’s life.
🛠️ How We Built It
We engineered MedSplit with a cutting-edge, modular architecture focused on speed, security, and scalability:
- Dynamic Frontend: Built with Next.js, delivering a fast, responsive interface that ensures an intuitive and engaging user experience.
- Powerful Backend: Implemented in Flask, seamlessly integrated with a high-speed JSON database for lightning-fast data retrieval and processing.
- Secure Core Functionality: Added robust authentication protocols, streamlined file/document sharing, and a custom payment flow — all supported by standard CRUD operations and RESTful APIs.
- Scalability & Evolution: Designed for growth, with planned transition to Django and MySQL for enterprise-grade scalability.
- AI Integration: Incorporated the Gemini API for translation, summarization, and chatbot functionality, enhancing accessibility and community support.
This stack allowed us to deliver a resilient MVP while laying a foundation for future features like wearable device integration, AI-powered reminders, and expanded financial services.
⚡ Challenges We Faced
- Scope vs. Time: Hackathons are short! Balancing ambition with MVP feasibility was tough.
- Integration hurdles: Ensuring seamless communication between Flask, MongoDB, and Auth0 while keeping the frontend responsive.
- Healthcare verification: Designing a doctor-approval system that was simple yet trustworthy.
- Track alignment: Clearly tying into Commure’s healthcare challenge, Capital One’s financial challenge, and Gemini’s AI challenge took deliberate planning.
🌍 Final Reflection
MedSplit isn’t just a hackathon project — it’s a vision of community-driven healthcare. We built it with the belief that no one should be denied treatment because of cost, and no medicine should be wasted when it could save another life.
Looking forward, we see MedSplit evolving into a platform that:
- Automated Calls & Reply to provide AI-powered reminders and instant information.
- Translation features to break down language barriers, making MedSplit accessible to all communities.
- Wearables Integration with healthcare devices like Dexcom for proactive, continuous care.
Our journey proved that small, thoughtful innovations can make a huge impact. Millions of heartbreaking stories exist where patients were denied medicine — but with MedSplit, there can be millions of solutions.
sion only, updates to come
Built With
- flask
- geminiapi
- github
- javascript
- mysql
- nextjs
- python
- react
- typescript

Log in or sign up for Devpost to join the conversation.