miniMDM is a minimal, self-hosted application for curating the records your business runs on — customers, products, locations, anything. Define your schema in YAML, edit through a clean web UI or REST API, and get versioning, audit trails and access control out of the box.
miniMDM ships the primitives that serious master-data work actually requires. No vendor upgrades, no per-seat licensing, no modules to buy separately.
Define every data object — fields, types, references, required flags — in a single YAML or JSON config file. Assign an owner and steward to each object for governance accountability. Version it in Git alongside the rest of your infrastructure.
Every edit writes a new version. Browse the complete history of any record, diff what changed, and revert to any prior state — no backup tape required.
Move records through draft → active → retired states without breaking API consumers. Edits create an isolated draft copy; publishing promotes it back to the golden master record. Retired records are excluded from default API responses but never deleted. Governance rules can enforce draft review before activation, prevent retirement of stable reference data, or require explicit approval for bulk imports. Webhook callbacks fire on every transition so downstream systems stay in sync.
Every golden record carries its origin — which source system it came from (ERP, CRM, warehouse, or any other upstream) and its ID in that system. Provenance is captured at import or creation, preserved in every history snapshot, and filterable from both the UI and API. Know not just what your master data says, but exactly where it came from.
What changed, who changed it, when, and why — searchable and filterable.
Four permission tiers — Viewer, Editor, Publisher, Admin — with per-schema grants for fine-grained control over who can read, write, or promote records.
CSV, TSV and JSON round-trips, with optional upsert-by-key.
Search within any object. Sortable columns and sorted dropdowns keep forms usable.
Cross-object references and parent-child relations, surfaced inline on the detail view.
Every object gets a documented endpoint and OpenAPI spec — automatically.
No marketing mock-ups — these are the real screens, captured from the live app.
01 Records list — sortable columns, full-text search and bulk actions on every object.
02 Detail view — fields grouped into sections with child records and back-references inline.
03 Version history — every edit stored forever, any version revertable in one click.
04 Audit log — a filterable trail of every action across every schema and user.
If your reference data has outgrown a spreadsheet but doesn't need a multi-year rollout, you're exactly who miniMDM is for.
| Shared spreadsheet | miniMDM | Enterprise MDM suite | |
|---|---|---|---|
| Time to first record | Minutes | Minutes | Weeks to months |
| Record-level versioning | — | ✓ | ✓ |
| Lifecycle states (draft / active / retired) | — | ✓ | ✓ |
| Full audit log (who / what / why) | — | ✓ | ✓ |
| Role & schema-level access control | Sheet-level | ✓ | ✓ |
| REST API + OpenAPI spec | — | ✓ | ✓ |
| Schema defined as code (Git-tracked) | — | ✓ | Usually UI-only |
| Self-hosted, no vendor lock-in | ✓ | ✓ | — |
| Cost | $0 (with caveats) | Free & MIT licensed | $$$ per seat, per year |
miniMDM is MIT licensed and lives entirely on your own infrastructure. Clone the repo, run it with Docker, read every line of the source.
Built on FastAPI, SQLAlchemy and PostgreSQL. Every change is tracked through Alembic migrations; every request gets a correlated log line. Deploy it on your laptop, a homelab VM, or behind your company's reverse proxy.