AI Use Case Risk Classification demo
Each new AI idea runs down one classification path — EU AI Act, GxP criticality, model risk, personal data and vendor terms — lit branch by branch as the intake is typed, each branch with its clause. 1:15
What the demo shows
An AI governance solution for risk teams that sorts new AI ideas under the EU AI Act and GxP.
As the sponsor types, agents read the description and the attached design, model card and vendor terms, and light the classification path: EU AI Act category, GxP criticality under draft Annex 22, FDA model risk, personal data and vendor terms — each branch with its reason and clause. The path ends in the company's review tier with reviewers routed; when an idea is not allowed as designed, it offers redesigns that re-light the path. The AI review board decides and signs, and the use case enters the AI inventory.
Transcript
- AI Use Case Risk Classification sorts new AI ideas for pharma risk teams under the EU AI Act and GxP.
- Today, one idea waits five weeks for four separate reviews.
- The sponsor describes the idea and the path lights as they type.
- Automated batch release by a generative model, critical GMP, high risk under Annex 22.
- Each branch gives its reason.
- Each reason opens its clause.
- Dana's deviation assistant fails, too.
- It would set the deviation class itself.
- One redesign, a QA reviewer decides, and the path relights as approvable.
- Dana signs, and it enters the AI inventory.
- Classified in minutes, each conclusion tied to its clause.
- Build on Botminds, agents, data, screens, and governance on one platform.
Transcribed from the film’s narration. Captions are on the player.