Help Me Damnit

Inspiration

Healthcare bureaucracy is a nightmare that leaves people fighting for their lives while drowning in paperwork. We built HelpMeDammit because everyone deserves an AI advocate who knows the system and won't back down.

What it does

HelpMeDammit is a multi-agent AI system that provides specialized healthcare advocacy. It intelligently routes users to Crisis Navigator (emergencies), Benefits Advocate (Medicare/insurance), or Appeal Fighter (claim denials) agents, with real Medicare benefits data integration.

How we built it

Built on Databricks with MLflow model serving, Unity Catalog, and Delta tables. Multi-agent architecture with intelligent triage routing. React frontend with glassmorphism UI deployed on Vercel. Integrated real Medicare benefits analysis data for accurate plan recommendations.

Challenges we ran into

Python environment conflicts with Databricks SDK, CORS policy blocking frontend-to-Databricks communication, Unity Catalog permissions issues, and Vercel serverless function module syntax compatibility problems.

Accomplishments that we're proud of

Created a fully functional end-to-end AI advocacy system in record time. Successfully integrated real Medicare data, deployed a production-ready model on Databricks, and solved complex CORS issues with a serverless proxy architecture.

What we learned

Databricks ecosystem integration, multi-agent AI architecture patterns, the complexity of healthcare data systems, and how to rapidly prototype and deploy AI applications with modern toolchains.

What's next for Help Me Damnit

Expand to insurance claim automation, add document parsing for medical records, integrate with healthcare provider APIs, and build a mobile app for on-the-go advocacy during medical emergencies.

Built With

• css3-databricks-platform:-mlflow-(model-training
• databricks
• databricks-model-serving
• glassmorphism
• javascript
• javascript/jsx
• mlflow
• python
• react
• registry
• unity-catalog
• vercel