Skip to content
Lorca Labs
Live demo
Evidence & custody

Confidential Model Registry

A lightweight model fingerprinting registry that content-hashes weights and records registrations with status — for verifying what was deployed, not for serving models.

The problem

When you need to verify which model version was registered with which weights hash, you need a registry that records the model ID, version, architecture, weights hash, and publisher — with a content hash for integrity — while being honest that this is a demo-grade registry, not a production model store.

What it does

01

Registers a model by modelId, version, weightsHash (64-char SHA-256 hex), and optional architecture/publisher.

02

Returns the full registration object with id, tenantId, status (active), and registeredAt.

03

Lists all registrations with content hashes, status, and timestamps.

Who it’s for

  • ML engineers verifying model provenance
  • MLops auditors checking registered weights
  • Compliance teams reviewing model registrations

In context

Complements Evidence Locker (anchored evidence), Supply Chain (artifact provenance), and the Consent Receipt Registry (content-hash integrity pattern).

Try it live in the studio

A no-signup, in-browser experiment against an isolated demo host. Load sample data, run the flow, and see the result — nothing is stored.