Case Study: Novo Nordisk achieves scalable pharmaceutical manufacturing optimization with Amazon Web Services

Novo Nordisk Uses ML for Computer Vision to Optimize Pharmaceutical Manufacturing on AWS

Novo Nordisk, a pharmaceutical manufacturer, needed a scalable and traceable way to automate quality-assurance tasks using computer vision and machine learning in a highly regulated environment. Using Amazon Web Services and Amazon SageMaker Pipelines, the company set out to streamline model development, deployment, and monitoring for its cartridge-counting process.

Amazon Web Services helped Novo Nordisk build an end-to-end ML pipeline on AWS that processes images, trains and tunes models, deploys them to edge devices, and monitors performance with tools like Amazon QuickSight and Amazon Timestream. The solution improved traceability, enabled anomaly detection for issues such as blocked cartridge boxes, and proved reusable when Novo Nordisk quickly adapted it to detect bacteria growth on agar plates with minimal changes.

Novo Nordisk

Ribera Codina

ML Engineer