AI Governance Institute
All governance templates →How do we build and maintain an AI model registry?

Implementation Kit

AI Model Registry Template and Entry Checklist

The registry as a governed source of truth, not a spreadsheet nobody updates. A field schema, a pre-deployment entry gate, a versioning policy that says what triggers a new record, and a retirement process.

Who this is for: The MLOps or governance owner standing up a model registry and wiring it to approval and monitoring workflows.

Download the kit (Markdown) ↓4 artifacts. Every table also copies as CSV.

1. Model registry data schema

Spreadsheet

The fields every entry must carry. Required fields gate deployment; optional fields add context.

Template

FieldRequiredDescriptionExample
model_idYesStable identifiermdl-fraud-scoring
versionYesSemantic version of this record3.2.0
ownerYesNamed accountable individual
use_caseYesWhat decision or output it produces
lifecycle_statusYesPreview / Testing / Production / Deprecated / Retired
risk_tierYesFrom the classification method
training_dataYesSources, cut-off, lawful basis
applicable_regulationsYesNamed regimes
eval_summaryYesLatest performance, and bias where people are affected, with dates
last_reviewYesDate of last governance review
upstream_dependenciesNoBase models, vendors, datasets
deployment_targetsNoWhere it runs

Worked example

FieldValue
model_idmdl-resume-screen
version2.4.0
ownerD. Portillo, Head of Talent
use_caseRanks applicants 1-5 for recruiter triage
lifecycle_statusProduction
risk_tierHigh
training_data5 yr hiring outcomes (internal), base model VendorLM-4; lawful basis legitimate interest with notice
applicable_regulationsEU AI Act Annex III, NYC LL144, GDPR
eval_summaryAccuracy 0.81 (2026-08); adverse-impact ratio 0.88 (2026-08), above the 0.80 line
last_review2026-08-30
upstream_dependenciesVendorLM-4 base model; Greenhouse ATS

Acceptance criteria

  • Every required field is defined with a data type and an allowed-values list where relevant.
  • The schema is the same one the deployment gate checks against.
  • Regulation and eval fields carry dates, so staleness is visible.

2. Registry entry pre-deployment checklist

Spreadsheet

The hard gate. No model reaches production without a complete entry.

Template

#CheckStatusBlocker if not met
1All required schema fields populatedY / NYes
2Risk tier assigned via the classification methodY / NYes
3Latest evaluation attached, dated within 90 daysY / NYes
4Human oversight design documented (High tier)Y / NYes for High
5Applicable regulations confirmed with LegalY / NYes
6Owner has accepted accountability in writingY / NYes
7Rollback plan recordedY / NYes

Worked example

#CheckStatusNote
1Required fields populatedY
2Risk tier assignedYHigh
3Evaluation within 90 daysY2026-08-30
4Human oversight design documentedPartialOverride exists; no gate on low scores. Blocker.
5Regulations confirmed with LegalYMemo 2026-03
6Owner accepted in writingY
7Rollback planYRevert to v2.3.1, 15 min
Gate decisionBlockedClose item 4 before deployment

Acceptance criteria

  • The checklist is enforced in the deployment pipeline, not run on the honor system.
  • A failed check blocks deployment unless a named approver records a time-boxed exception.

3. Model versioning policy

Document

Says which changes create a new registry version and which of those need a fresh governance review.

Template

One page. Attach to the schema.

A new version record is created when any of these change:

  • Model weights (retrain, fine-tune, base-model swap)
  • The prompt or system instructions, for prompt-based systems
  • Inference parameters that affect output (temperature, thresholds, retrieval config)
  • The training or retrieval data set
  • The intended use case or the population affected

A new version also triggers a governance review when:

  • The risk tier could change
  • The change affects a High-tier system
  • Evaluation shows a material shift in performance or bias
  • A new regulation now applies

Every version record carries: version number, date, author, changelog, evaluation delta, reviewer (if triggered).

Worked example

Change: swapped base model from VendorLM-3 to VendorLM-4 on the resume screener.

  • New version: 2.4.0 (weights changed).
  • Governance review triggered: yes (High-tier system; eval delta).
  • Changelog: base model upgrade; re-ran adverse-impact eval, ratio moved 0.85 to 0.88.
  • Reviewer: R. Nkemelu, 2026-08-30. Outcome: approved, monitoring cadence unchanged.

Acceptance criteria

  • The list of change types that create a version is specific enough that two engineers would agree.
  • Every production model in the registry has at least one version record with a changelog.
  • Review-triggering changes have a linked reviewer and outcome.

4. Model retirement process

Spreadsheet

Decommissioning steps so a retired model does not linger in code, storage, or documentation.

Template

StepOwnerDoneEvidence
Confirm no production traffic and no scheduled jobs call the modelY / N
Migrate or sunset dependent systemsY / N
Delete or archive model weights per the data policyY / N
Archive documentation, evaluations, and version historyY / N
Set registry lifecycle_status to Retired with a dateY / N
Notify stakeholders and update the AI inventoryY / N
Owner sign-offY / N

Worked example

StepOwnerDoneEvidence
Confirm no traffic or jobsMLOpsY30-day traffic log, zero calls
Migrate dependentsMLOpsYLegacy chatbot moved to support copilot
Delete or archive weightsMLOpsYWeights archived to cold storage, prod copy deleted
Archive documentationAI Gov LeadYMoved to governance archive, 7-yr retention
Set status to RetiredAI Gov LeadY2026-09-15
Notify and update inventoryAI Gov LeadYInventory row marked retired
Owner sign-offHead of SupportY2026-09-15

Acceptance criteria

  • Retirement is not complete until every step has evidence and an owner sign-off.
  • The registry and the AI inventory both show the model as retired with a date.

Governance controls this kit produces evidence for

Completing the artifacts above gives you a head start on the evidence requirements for these controls.

CHM-001
CHM-001

The schema and versioning policy are the model version control standard, with a version record per change.

CHM-004
CHM-004

Each version record's changelog and evaluation delta are the model change documentation.

CHM-005
CHM-005

The retirement process is the model deprecation procedure with evidence per step.

MGV-002
MGV-002

The pre-deployment entry checklist is the intake and approval gate.

SCT-009
SCT-009

The populated registry serves as the algorithm register for production models.

This kit backs one playbook. Read the full guidance for the reasoning behind each artifact.

Decide what to implement next

Assess your governance gaps, then create an action plan with owners and target dates. Build and export without an account; sign in when you want to save your plan.

Start the AI governance assessment →