One capture path
UI, API, batch jobs, and ad hoc SQL all hit the same database. Row-level capture records INSERT, UPDATE, and DELETE regardless of channel — no per-feature audit rewrites.
Temporal audit infrastructure
Database history you can prove, inspect, and reconstruct.
Trustworthy row-level history from day one — every change path through your database, tamper-evident and inspectable with SQL.
How it works
Foundation is the product
Append-only history, hash-chain verification, and point-in-time reconstruction through SQL. Compliance, forensics, and debugging without waiting for a UI.
Install early
Run migrations when PostgreSQL is provisioned; register tables as you introduce them. Set-and-forget after install — history from the first meaningful data.
Use cases
- Debugging and release confidence — reconstruct what changed before and after a deploy.
- Compliance and evidence — inspectable, verifiable change history for audits.
- Investigation and accountability — who changed what, when, with before/after state.
- ML training data provenance — reconstruct datasets at a point in time with integrity checks on the audit trail. Data provenance for ML governance — not an “AI audit” product.
Foundation and Application layer
Foundation
Database capture, tamper-evident store, SQL inspection and verification. Complete product value today — Foundation release is SQL-first and in active development.
Application layer
Optional UIs, APIs, and workflows that surface the same history. Additive — not required to realise Foundation value. On the roadmap after Foundation is validated.
Get started
PostgreSQL Audit (Foundation release) lives in the monorepo: migrations, architecture, and product decisions are documented there.