Inspiration

Medication errors and drug interactions are a major cause of preventable harm worldwide. Many people take multiple medicines daily without fully understanding the risks involved. Existing medical data is often scattered, complex, and inaccessible to non-experts.

This inspired us to build medSafe - a system that simplifies drug safety and makes it accessible to everyone.


What it does

medSafe helps users identify potential drug risks in a simple and reliable way.

  • Extracts medicine names from prescription images or manual input
  • Converts brand names into standardized generic names
  • Detects drug interactions and contraindications
  • Presents results in clear, user-friendly language

It transforms complex medical data into actionable safety insights.


How we built it

We developed medSafe using a full-stack architecture:

  • Frontend: React.js (TypeScript)
  • Backend: Node.js
  • Database: MongoDB

Core integrations:

  • RxNorm API → Drug normalization
  • OpenFDA API → Drug safety data
  • OpenAI API → Structured data extraction

Workflow logic:

  1. Input (image/manual)
  2. Extraction
  3. User verification
  4. Normalization
  5. Interaction analysis
  6. Data storage
  7. Output display

Challenges we ran into

  • Handling noisy prescription images
  • Managing multiple drug name variations
  • Dealing with incomplete OpenFDA data
  • Dependency on external APIs
  • Converting unstructured medical text into structured insights

Accomplishments that we're proud of

  • Built a complete end-to-end pipeline
  • Implemented RxNorm-based normalization
  • Enabled bidirectional interaction checking (A ↔ B)
  • Designed a clean, user-friendly interface
  • Reduced API calls using database caching

What we learned

  • Integration of real-world healthcare APIs
  • Importance of data normalization
  • Handling unstructured data using AI
  • Building systems that balance accuracy, usability, and scalability
  • Designing for real-world healthcare impact

What's next for medSafe

  • Expand to global drug datasets
  • Support multilingual access for users across different regions
  • Integrate speech-to-text input to assist differently-abled users and improve accessibility
  • Enable healthcare provider integration

Built With

Share this project:

Updates