Matruno Documentation

What Matruno Is#

Matruno is a Legal Matter Operations platform. It keeps matters, participants, documents and evidence, legal workflow state, human legal review, approvals, generated legal documents, agent activity, notes, timelines, and execution logs tied to one workspace-scoped operating surface.

It is built for law firms, in-house legal departments, and legal operations teams that need to move confidential legal work through clear steps: legal intake, create a matter, attach evidence, run a legal workflow, pause for attorney review when needed, generate a legal deliverable, and reconstruct what happened later.

Matruno is not an autonomous lawyer, a legal-advice chatbot, a generic CRM, or a simple document generator. The product center is governed AI workflows for legal matters with human legal oversight.

Core Operating Model#

1. Set up a workspace and create the first admin/operator accounts.
2. Model a matter with the fields, participants, documents, and lifecycle state required for the work.
3. Define a workflow with ordered steps, dependencies, branching, terminal steps, and optional template bindings.
4. Add human review where policy requires approval or structured feedback.
5. Upload source documents, manage document templates, and generate fixed-layout or flowing artifacts.
6. Observe execution through matter timelines, agent logs, task status, Temporal worker health, and OpenTelemetry signals.

Quick references#

Architecture at a Glance#

The repository implements a FastAPI backend, Next.js frontend, optional Next.js admin panel, PostgreSQL source of truth, Redis cache/coordination layer, Temporal server and workers, local artifact storage, OpenTelemetry collector, and Grafana dashboard/rule bundle.

The default local stack is managed by root docker-compose.yml:

Recommended Next Steps#

Start with local development, then complete workspace setup. After that, create a narrow first matter, import or define one legal workflow, add documents through Documents, and use Troubleshooting if the backend, worker, Temporal, PostgreSQL, Redis, or browser app does not behave as expected.