Vero

1.3 million people are harmed annually by medication errors in the US. Most existing tools are binary, they say yes or no without context. Vero aims to solve this problem with HPC-powered analysis.

Comment 1

Zoom Link: https://tamu.zoom.us/j/4978426445?omn=95240024708

Inspiration

Angel's mother, Mrs. Figueroa, suffered serious medical complications as a result of an undetected drug interaction between warfarin and aspirin.

What it does

Vero uses Monte Carlo simulation and Claude AI verification to provide the first truly personalized, probabilistic drug interaction analysis

How we built it

Vero uses a FastAPI async Python backend running HPC simulations (Monte Carlo, multi-stage analysis), connected to a Next.js frontend for visualization. The system combines computational analysis with AI verification to provide personalized risk assessments.

Challenges we ran into

Integrating and normalizing data from multiple sources (DrugBank, FAERS, interaction databases) with inconsistent formats

Translating complex probabilistic results into clear, actionable insights for healthcare providers

Designing an ambitious frontend in Figma.

Accomplishments that we're proud of

Helping those close to us, while using tools that allow us to learn and develop our careers.

What we learned

  • how to design a job queue system that tracks progress across multiple simulation stages while maintaining state consistency
  • Async FastAPI architecture for long-running simulations ## What's next for Vero Expanded drug database Mobile app for patients Work with providers like Epic Uploading previous medical documents

Built With

Share this project:

Updates