Case study
AI medical-imaging app
Medical product using an ML pipeline to analyze joint X-ray imagery, delivered directly inside the client's engineering workflow.
45+ tasks • 3 contributors • 4 departments
Challenge
Backend work had to fit a client-owned ML codebase and GitHub issue flow.
Outcome
Backend and ML-pipeline delivery coordinated issue-by-issue through the client's GitHub flow: doctor/patient auth, imaging history, and pipeline orchestration.
Team mix: 2 backend, 1 PM/BA.
My contribution
Coordinated Celery and RabbitMQ pipeline orchestration, an Azure database migration, API and schema changes, and auth/verification flows in step with the client's in-house ML team.
What shipped
- AI imaging workflow delivered inside the client's ML codebase
- Doctor/patient auth and imaging history
- Pipeline orchestration on Celery/RabbitMQ, Azure database migration
Stack
- Python
- Celery
- RabbitMQ
- Azure
- PostgreSQL
- ML pipeline integration
- GitHub