Skip to content

Is Facet right for you?

Decide before you install. This page gives you the fits, the misfits, and the business cases where Facet is the answer.

Facet works well when most of these hold:

Fit What it gives you
The business side builds its own apps No waiting on a development queue. A business person can describe the app to Claude Code, Claude Opus, or another coding agent, and ship it in hours.
You want agentic speed without losing control Facet gives the agent a governed surface: it acts under one role you pin, and the platform checks every call, whatever wrote the code.
The records must outlive the apps A customer stays a customer, whatever tool edits it.
People see different rows and fields An auditor may ask who changed what.
You are a small or mid-sized organization and want to self-host The data stays in your PostgreSQL, on your machines.

Facet replaces the classical stack under the app — the bespoke backend, the auth plumbing, the audit scripts. It does not replace the app; it makes the app small enough to throw away.

Facet pays back best in these five situations:

Situation What it gives you
Sovereignty of your data The data stays in your PostgreSQL, on your machines. Nothing leaves unless you send it.
Auditability and accountability is mandatory Every write lands in an append-only event log, with who, when, and what. This covers regulated industries such as healthcare and food.
Free code you can deploy or adapt You hold the code. Deploy it, fork it, or change it to fit your organization.
Limited access to software developers Business people describe the app to a coding agent and ship it. The platform carries auth, rights, and audit, so the app stays small.
A business needing custom-made domains and applications Describe the domain in a draft, and Facet gives you the records, lists, rights, and history for it.

Facet is opinionated on purpose: one model of domains, doctypes, roles, and history, applied to every write. These workloads need a lower level of control. Use custom-made, less opinionated software for them — a plain database and hand-built services:

Workload Why it does not fit Use instead
Heavy data load Every write is validated, logged, and projected. That costs time and storage per row. A plain PostgreSQL schema, tuned for the load
Statistics and analytics Lists and filters serve daily operations, not aggregate scans over very large row counts. An analytics engine, fed by exports
B2C products Facet serves the people of one organization. Accounts come from invites and SSO — there is no public self-service sign-up. A custom-built consumer stack
Multi-tenancy One deployment serves one organization. Facet has no walls between separate customer companies. A purpose-built multi-tenant design
Time-series entries High-frequency timestamped rows want append-only storage with retention pruning, not per-row history. A time-series database
A business-specific IT architecture Facet imposes its model as-is: domains, doctypes, roles, and the event log. It does not reshape around a house architecture. Custom services around your own architecture
Wanting a hosted service Facet is self-hosted software; you run the servers. A hosted platform

Run the stack, click through it, and build one small domain of your own:

  1. Run the stack:

    Terminal window
    git clone https://github.com/wearefacet/facet.git
    cd facet
    cp docker/.env.example docker/.env # fill in hostname, SMTP, secrets
    docker compose -f docker/compose.prod.yml up -d --build
    docker compose -f docker/compose.prod.yml exec backend facet seed:admin ops@your-company.tld
  2. Sign in with the printed one-time password.

  3. Click through the workspace: lists, filters, document history.

  4. Build one small domain of your own — Domains, drafts and domain editors.

If step 4 took an hour and most of the good fits hold for you, Facet fits.