Reading pipeline health at machine speed

A Healthcare & Life Sciences business working with Cloud Combinator on AWS. The client is anonymised at their request.

The challenge had three focus areas, each about turning an expert visual judgement into a reliable automated service.

Cloud Combinator delivered the work through its Cloud Accelerator Program, a staged path that moves a client from idea to a running solution in their own AWS account while upskilling their team.

The engagement delivered a working, automated classification pipeline in the client's own AWS account: send an image to the API, and Amazon Rekognition returns a zero-to-seven pipeline-health score as structured JSON. Acceptance confirmed the Rekognition and SageMaker platform as a suitable environment for the client to continue developing the product.

• Automated health scoring — A Rekognition model trained on the client's imagery classifies pipeline health on the familiar zero-to-seven scale, removing the manual inspection step.
• Scales with the imagery — Built on managed AWS services, the pipeline uses automatic scaling to handle a heavy, continuous stream of camera images.
• Simple to integrate — Images go in over an API and results come back as JSON, making the service easy to consume from the client's own systems.
• Best practice baked in — Secure S3 storage and permissions plus a Well-Architected Framework Review give the solution a sound operational footing.
• Handover — The project was returned to the client with the solution in their AWS account and the internal technical lead upskilled to continue model development in SageMaker.

With the platform proven and their tech lead upskilled, the client is positioned to extend and retrain the model across more pipelines and cameras, scaling automated health monitoring across their estate.