Integrations

Out of the box, Teampilot can chat, search the web, and process the files you upload. Integrations go further: they let the AI reach into your databases, talk to external services like Zapier and Linear, run built-in AI functions, call your own HTTP endpoints, and expose Teampilot itself through an API. Together they turn chat into a hub that can read your data and trigger actions in the tools you already use.

You manage integrations under Settings > Integrations. Most integrations are team-wide and admin-managed — once an admin connects a service, everyone on the team can use it from chat and in Launchpads. A few integrations can also be configured personally so they only apply to your own chats.

Connect a database and the AI can query your data directly and answer questions from it — no exporting or copy-pasting. This is ideal for asking plain-language questions about your business data and getting answers, tables, and charts back in chat.

Teampilot supports the following database systems:

• PostgreSQL
• MySQL
• BigQuery
• Neo4j
• MSSQL (Microsoft SQL Server)

1. 1Open Settings > Integrations and add a new database connection (admins only).
2. 2Choose the database type and enter the connection details (host, port, database name, and credentials).
3. 3Save the connection so it becomes available to the team.
4. 4In a chat, ask a question about your data — the AI uses the connection to query it and answers from the results.

Connect Teampilot to Zapier to trigger automated workflows across thousands of apps. With Zapier you can hook chat into CRMs, spreadsheets, messaging tools, calendars, and much more, so the AI can kick off actions in services Teampilot does not integrate with directly.

Zapier is currently in beta. Behaviour and available actions may change while the integration matures.

The Linear integration gives the AI access to your Linear issues, so it can help with project management tasks — summarising issues, drafting updates, and answering questions about your backlog directly in chat.

Linear is currently in beta. Expect ongoing improvements to the available actions and data it can reach.

Teampilot ships with 30+ AI functions — built-in capabilities the AI can call on its own when a task needs them. In Smart Mode the AI decides which functions to use, so you usually just ask in plain language and let it pick the right tool. See Chat for more on Smart Mode and tools.

Built-in functions cover areas such as:

• Data analysis — work with numbers, tables, and structured data.
• Web access — search the web and fetch pages for up-to-date information.
• Document processing — read and extract content from uploaded files.
• Code execution — run code to compute results.
• Image processing — analyse and generate images.

Custom HTTP functions let you connect the AI to your own APIs. You define a function that calls any HTTP endpoint, and the AI can then invoke it as part of a conversation — for example to look something up in an internal system or to trigger an action in your own service.

1. 1Open Settings > Integrations and create a new custom HTTP function (admins only).
2. 2Give it a clear name and description so the AI knows when to use it.
3. 3Configure the endpoint URL, method, headers, and any parameters the function accepts.
4. 4Save it — the function becomes available to the AI in chat and can be added to Launchpads.

API keys give you programmatic access to Teampilot, so you can integrate its AI assistants into your own applications and scripts. Each request uses an API key together with your team ID to authenticate.

API keys are admin-managed: admins create and revoke keys under Settings > Integrations. Treat keys like passwords — anyone holding a key can act against your team, so store them securely and revoke any key that may have been exposed.

The API is currently in beta and may change as it stabilises.