Datachef documentation

Datachef is a full-stack data platform for revenue teams. It syncs your business systems (Salesforce, Marketo, Google Analytics 4 and your own databases) into one place, lets you query them together by asking a question in a sentence or writing SQL, and keeps your team's definitions as shared, versioned models. You don't need to set up a warehouse, an ETL tool or a BI tool.

Where to start

• New to Datachef? Follow the quickstart. It takes about five minutes with the sample dataset, and you don't need credentials.
• Connecting a source? Read connections first, then the page for your source: Salesforce, Marketo, Google Analytics 4 or PostgreSQL and MySQL.
• Writing queries? The SQL editor page covers the dialect, how tables are named and the limits.
• Using Claude? See the Claude connector.
• Setting up your team? Start with people and roles.

How Datachef fits together

1. Connect. Each source becomes a connection. Its objects become tables you can query, such as salesforce.opportunity.
2. Sync. Each table syncs on its own schedule. It can be a full refresh, or incremental on a cursor column like SystemModstamp.
3. Explore. Ask a question on the Agent page, or write SQL in the SQL editor. Every result comes with a chart.
4. Model. Save a query as a model: a named, versioned definition that everyone in the workspace (and any AI agent you connect) queries by name.
5. Share. Publish an analysis tab to share it with the workspace, read-only.

Key concepts defines each of these terms.

Getting help

Every plan includes a dedicated solution engineer. If something here is unclear or wrong, contact us and we'll fix the page.