Book a Demo

SLayer v0.5 is live — the open-source semantic layer built for AI agents. View on GitHub →

The governed semantic layer your agents run on

Motley turns your warehouse into data agents can query safely. Define every metric once, expose it over MCP, and any agent gets governed SQL and the same answer every time.

What Motley connects to and creates

From your warehouse to answers your agents can trust

Motley connects your business data to AI agents through a trusted semantic layer — helping them query approved metrics, follow business logic, and generate reports, charts, documents, and presentations.

Connect your data

Bring in the databases and warehouses your team already uses. SLayer introspects the schema and drafts the semantic model automatically.

Govern it in the layer

Define metrics, dimensions, and business rules once. Agents query those, never the raw schema, so the numbers are right every time.

Let any agent put it to work

Claude, Cursor, Codex, or your own agent queries over MCP and turns governed results into reports, documents, and reusable templates.

Semantic Layer

One definition of every metric, enforced everywhere

Configure and audit your semantic layer from your preferred coding agent

Manage your database connectivity: Motley supports SQL, Postgres databases, DuckDB, Clikhouse and most major warehouses

Define and manage your data model with automated introspection and MCP-based coding-agent extensions. Motley’s API and MCP server expose a superset of the SLayer interface, adding production controls for orchestration, generation, versioning, and auditability.

Visualise your semantic layer: auditability and versioning in a simple UI

Output generation

Documents built on governed queries

Most real questions need more than one query. A document combines queries, tables, charts, and written analysis into one governed artifact, not pasted together by hand.

Multi-object documents

Most data enquiries require more than one query to be answered; documents combine several objects: queries, tables, charts, and text analysis generated from query results

MCP-triggered generation

Generation runs through Motley's MCP server, with every metric validated against the semantic layer before it lands.

Versioned & auditable

Documents are stored, versioned, and fully auditable in Motley's UI.

Agent-ready recall

Agents can recall any document as context for the next question.

Masters

Define the report once. Run it forever.

Masters are reusable templates. They let agents generate the same business document over and over, with fresh data and variables each time, so recurring analysis stays consistent without being rebuilt.

Recurring analysis

Recurring analysis, whether a scheduled report or an agent running a workflow, needs to stay consistent. A Master runs the same analysis again without rebuilding it each time.

Deterministic generation

Masters resolve deterministically from key variables like time period or customer name, so the same template produces a fresh, accurate output every time with no manual setup.

MCP resolution

Masters are stored in Motley with full version history and resolve directly through the MCP server, so any connected agent can trigger generation.

Audit & versioning

Every Master is stored, versioned, and fully auditable in Motley's UI, so you can track how outputs change over time.

Start building with an agent-native semantic layer

Define business logic once.
Query consistently across agents, apps, and Python workflows