Use case · AI governance
Govern AI like the regulator is watching.
The EU AI Act asks a simple question with a hard answer: can you show what your model learned from, and what it did? On Polnor, lineage, audit and the registry answer it by default, for training and inference alike.
Why Polnor
Why us for this.
Lineage from model to data
Every training run records the tables it read. A model version traces back to the exact data it saw.
Audit on both sides
Training runs and serving calls land in the same audit trail, exportable to your own S3.
A registry with stages
Versioned models move through Staging and Production deliberately, with back-links to the endpoints serving them.
Evidence in your jurisdiction
Artifacts, logs and the audit base live on your bucket, under your keys, in the EU.
What you can build
AI Act readiness
Build the documentation trail (data, training, evaluation, deployment) as a by-product of working.
Model risk reviews
Give risk and compliance a queryable view of what runs where, on what data.
Incident forensics
When a model misbehaves, walk the lineage back to the run, the code and the rows.
Set a course
Request a
demo.
See your own Iceberg tables, warehouses and notebooks running on your European cloud, usually within a week.
hello@polnor.net · OVHcloud GRA9, France 🇫🇷