Healthcare
Wasfa
An AI assistant that reads handwritten medical prescriptions
- PyTorch
- OCR
- FastAPI
- React
The problem it solves
Dispensing errors caused by misreading handwritten prescriptions are a documented global problem, and their regional cost is measured in real incidents, not just theoretical numbers.
The solution
A computer vision model trained on Arabic medical handwriting converts a prescription image into structured text, matches it against the registered medicines database, and flags any ambiguity to the pharmacist instead of guessing.
Impact
It began as a capstone project, moved into an operational pilot with a pharmacy chain, and is today a registered company serving more than 40 branches.
Awards & recognition
Best Capstone Project Award 2025
First place, Digital Health Hackathon
The team
Maram Alshehri
A former pharmacist turned technology founder. She builds systems that reduce dispensing errors, and believes the best technical solution comes from whoever lived the problem.
- PyTorch
- OCR
- Computer Vision
- FastAPI
Omar Alghamdi
Works on visual analytics systems that are private by design and store no personally identifying data whatsoever.
- Computer Vision
- Tracking
- Python
- Streamlit
Related content
- Scikit-learn
- Pandas
- Streamlit
They were exactly where you are… look where they are now.
Membership is automatic for every graduate who completes a program — no application, no fee.