Welcome to Motley¶

What Motley does¶

Motley turns your warehouse into a governed, agent-ready analytics platform, so every report your agents generate is auditable, repeatable, and built on a single definition of every metric.

How Motley is different¶

Instead of exposing raw tables, Motley builds a Semantic Layer that sits between your database and AI agents, providing them with a simple interface to generate queries in a natural way. Our semantic layer is specifically built for AI agents that need dynamic, predictable, and structured access to analytics logic. This makes queries more reliable, reusable, and easier to reason about.

The semantic layer you define once improves every document generated from that source. It also makes documents consistent across time and users: the same metric always means the same thing.

Architecture¶

Motley architecture: data flows from a source through the semantic layer to an AI agent, producing a Motley document

Data Sources: your databases (MySQL, PostgreSQL) and files (CSV, Excel) connected to Motley. Motley queries them live at document generation time so documents always reflect current data.

Semantic Layer: the structured model that sits between Motley and your data, generated automatically on ingestion. You can refine it with an agent or in our UI.

Reporting engine: the runtime on top. Compose queries, tables, charts, and narrative into versioned Markdown documents your team can forward, diff, audit, and regenerate on demand.

Agent: connected to Motley via the MCP connector, it reasons about your data through the semantic layer and assembles documents. Our Claude plugin is the easiest way to try it.

Documents: every document is fully auditable in the Motley UI.

Get started¶

→ Quickstart with Jaffle Shop demo data — generate your first document in under 5 minutes, no setup required

→ Quickstart with your own data — upload CSV/Excel or connect a database to generate documents from your real data