FastDataGov

FastDataGov is an open-source, business-friendly data governance platform built with FastHTML, PostgreSQL, and Python. It provides a unified catalog, lineage, data quality, stewardship, ownership, certification, and access-request experience across connected data platforms.

The platform is deliberately vendor-neutral. Snowflake is the first fully implemented adapter; Microsoft Fabric and Databricks implement the same public contract and can be enabled without redesigning the core.

What works

• Searchable catalog with platform, domain, trust, owner, and quality filters.
• Business glossary and asset-to-term relationships.
• Versioned business/field metadata, native and business tags, sensitivity, domain hierarchy and reusable data products.
• Asset detail pages with ownership, stewardship, sensitivity, freshness, quality, lineage, access guidance, and audit context.
• Interactive table- and column-aware lineage with impact views.
• Manual, dbt and generic lineage enrichment with explicit evidence labels.
• Versioned quality rules, run history, scores, alerts, and remediation work.
• Stewardship queues for quality remediation, certification, metadata enrichment, access requests, and ownership attestation.
• Adapter health, capability manifests, incremental sync cursors, and job runs.
• Configurable workflow due dates/approvals, comments/history, renewable asset and product certification, scoped accountability and notification outbox.
• Source-grant discovery with explicit identity-to-platform principal aliases.
• Microsoft Entra ID or Google production authentication and safe local developer auth.
• Paginated JSON API, directional impact API, dbt/generic lineage imports and source-filtered catalog CSV export.
• Deterministic demo data so the complete product is useful without cloud credentials.

Quick start

python -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt
# For live Snowflake work: pip install -e '.[dev,snowflake]'
cp .env.example .env
python main.py

Open http://localhost:5062. The checked-in example configuration uses local developer authentication and the deterministic demo repository. Sign in with any valid email address; no password or external identity call is made in this mode.

PostgreSQL mode

export DB_URL=postgresql://fastdatagov:fastdatagov@localhost:5432/fastdatagov
export DB_SCHEMA=fast_datagov
export REPOSITORY_MODE=postgres
python -m fastdatagov.db.migrate
python -m fastdatagov.synthetic.seed
python main.py

All database objects live in the fast_datagov schema. Migrations are append-only and recorded in fast_datagov.schema_migrations.

Public product pages

The public landing page explains FastDataGov's operating model and shipped capabilities. /features gives explicit availability for every core feature and priority adapter; /compare provides a dated, source-linked comparison with OpenMetadata, DataHub, Atlan, Collibra, and Alation. Comparison claims are deliberately limited to official vendor documentation and clearly separate software licensing from implementation and infrastructure costs.

Microsoft Entra ID

Set AUTH_MODE=entra and configure the variables documented in .env.example. Register /auth/callback as the HTTPS redirect URI. FastDataGov requests only OpenID identity scopes and stores no identity provider access tokens in application tables.

Google OpenID Connect

Set AUTH_MODE=google, configure GOOGLE_CLIENT_ID, GOOGLE_CLIENT_SECRET, and the access-list variables documented in .env.example, and register /auth/google/callback as the exact HTTPS redirect URI. State is validated by the OAuth client, only openid email profile is requested, and the returned email must be verified and permitted. Use GOVERNANCE_ADMIN_EMAILS to bootstrap the first governance administrators.

Commands

pytest -q                         # offline unit and application tests
python -m fastdatagov.db.migrate # apply pending PostgreSQL migrations
python -m fastdatagov.jobs.worker
python -m fastdatagov.jobs.scheduler

FastSME Coolify deployment

FastDataGov is catalogued by the sibling FastDevOps control plane. With deployment credentials in the ignored mode-0600 .env, use:

python scripts/coolify.py validate
python scripts/coolify.py doctor
python scripts/coolify.py status
python scripts/coolify.py env --sync --yes
python scripts/coolify.py deploy --yes

The declared production target is https://datagov.fastsme.com, port 5062, PostgreSQL repository mode, /healthz, and Google OpenID Connect at /auth/google/callback. Secret values remain in ignored files and Coolify.

See architecture, adapter development, API, implementation, pilot, operations, training and transition, capability matrix, security, and contributing. The maintained functional release gate is docs/COMPLETION_MATRIX.md.

License

MIT — see LICENSE.