Govern the full model lifecycle inside PostgreSQL

Model registry & metadata

Every model version is tracked in the NeurondB registry table with immutable metadata. Store provenance, signatures, and deployment configuration in JSONB to integrate with your compliance controls.

Register a version

SELECT neurondb_register_model(
  name        => 'reranker-cross-encoder',
  version     => '2.1.0',
  storage_url => 's3://models/neurondb/reranker-cross-encoder-2.1.0.onnx',
  runtime     => 'onnx',
  device      => 'auto',
  metadata    => jsonb_build_object(
    'owner', 'ml-platform',
    'git_commit', 'b4c5d9f',
    'trained_at', CURRENT_TIMESTAMP
  )
);

Inspect registry

SELECT name,
       version,
       metadata ->> 'owner'     AS owner,
       metadata ->> 'git_commit' AS git_commit,
       created_at,
       status
FROM   neurondb_model_registry
ORDER  BY created_at DESC;

Rollout controls & staged environments

Promote models between dev, staging, and production directly in SQL. NeurondB stores active deployment slots and supports canary percentages for gradual rollouts.

Promote model

SELECT neurondb_promote_model(
  name          => 'reranker-cross-encoder',
  source_stage  => 'staging',
  target_stage  => 'production',
  canary_weight => 0.25
);

Active deployments

SELECT stage,
       name,
       version,
       canary_weight,
       created_at
FROM   neurondb_deployments
WHERE  status = 'active'
ORDER  BY stage, created_at DESC;

Performance monitoring

Track model performance metrics and set up alerts for degradation.

SELECT 
  model_name,
  version,
  AVG(latency_ms) as avg_latency,
  PERCENTILE_CONT(0.95) WITHIN GROUP (ORDER BY latency_ms) as p95_latency,
  COUNT(*) as request_count
FROM neurondb_model_metrics
WHERE created_at > NOW() - INTERVAL '1 hour'
GROUP BY model_name, version;

Next Steps

• Inference Runtime - Deploy models
• Monitoring & QA - Track model performance
• Feature Store - Manage features