Gatlas/docs

Getting Started

Everything you need to go from zero to querying your portfolio data with AI.

What is Gatlas

Gatlas is a data infrastructure platform for holding companies and PE firms. It syncs data from your portfolio companies — Shopify, Stripe, QuickBooks, and more — into a unified data warehouse, then exposes it all to AI assistants via MCP. One place to connect, transform, and query data across every entity in your portfolio.

The Setup Path

1

Create your organization

Sign up and name your holding company or fund. This is your top-level account — everything else (entities, sources, views, API keys) lives under it.

2

Add entities

Entities are your portfolio companies. Each entity gets its own database schema and can have its own set of data sources. Navigate to Entities and click Add Entity to create one.

3

Connect data sources

For each entity, connect their tools — Shopify, Stripe, QuickBooks, and others. Most connectors use OAuth, so you just click Connect and authorize. A few require an API key instead. See the Connecting Sources guide for details on each connector.

4

Run your first sync

Once a source is connected, trigger a sync from the Sources page. Gatlas pulls all available data into your warehouse. After the first sync, subsequent syncs run automatically on schedule — you only need to trigger it manually the first time.

5

Create views

Views are saved SQL queries that transform raw synced data into useful, queryable tables. Use the SQL Workspace to write queries, then save them as views. Add AI context and column descriptions to make them understandable to AI assistants. See Views & SQL.

6

Share MCP access

Generate an API key and scope it to specific entities or views. Share the key with whoever needs access — they connect their AI assistant (Claude, Cursor, or any MCP-compatible client) and can start querying data immediately. See MCP & API Keys.

What's next

That covers the golden path. For deeper dives into each part of the platform, check out these guides:

  • Connecting Sources — auth types, connector-specific notes, delegated setup links
  • Views & SQL — writing queries, AI context, column docs, publishing views
  • MCP & API Keys — generating keys, scoping access, connecting AI clients