How it works
From submission to validated deployment
Every use case in AIH Lab travels a structured pipeline. Each stage builds trust, evidence, and replicability.
Stage 1
Submit
A use case is filed through a structured form built on the use case families schema, so every record starts comparable. It ends with four questions — what worked, what did not, why it was stopped or changed, and what the next hospital would need. For now, during a validation phase, it is filed together with us in a guided online session: ask for a demo.
Stage 2
Curate
The AIH Lab team confirms completeness, attaches the instance to its family, and tags duplicates. AI assists; humans decide.
Stage 3
Mature
Clinical, technical, ethical, and operational reviewers each cover their dimension. The output is a documented review trail per instance.
Stage 4
Validate
Real-world performance is confirmed and packaged as an Assurance Pack — evidence, oversight model, and operating conditions in one place.
Stage 5
Scale
Validated packs are formatted for replication: the next hospital starts with the same structure, not a blank page.
The role of AI
AI assists three stages: completeness checks at Submit, family suggestion at Curate, and synthesis at the Hub. Every decision is recorded; every output is labelled. A human is on the loop at every gate.