Bruin Cloud onboarding

AI Data Analyst - first run

A short walkthrough of the AI track. Watch the video, then work through the checklist - one required step, the rest are suggestions.

RequiredYou can't query your data without this.

  1. Connect your data

    Add a connection so the agent can query your warehouse. Supports BigQuery, Postgres, MySQL, SQL Server, Snowflake, Databricks, and Redshift. We recommend creating with validation so Bruin can confirm the credentials work.

    Don't have direct access to the warehouse? Invite a teammate from your data team to your Bruin Cloud workspace and have them set up the connection.

    Manage Connections
Recommended nextThe fastest way to get value out of the agent.

  1. Try the AI chat

    Ask a question in plain English and get back charts, tables, or numbers. A good first prompt is to ask the agent what data it has access to. The agent can also export CSVs and PDF reports inline.

    Chat with an AI Agent

  2. Connect chat integrations

    Bring the agent into Slack, Teams, Discord, or WhatsApp so your team can ask questions where they already work.

    Configure AI Agents

  3. Build a dashboard

    Pin chat answers as widgets your team can re-open without asking again. Prompt the agent to assemble charts and filters, then publish and share.

    Build Dashboards with AI

  4. Schedule an agent

    Run the agent on a cadence. Daily or weekly reports land in chat ready to read - no one has to remember to ask.

    Scheduled Agents
AdvancedImprove accuracy by giving the agent your team's context.

  1. Add a context layer from a Git repo

    Connect a repo so the agent learns your team's vocabulary and metric definitions from your dbt or Bruin semantic layer.

  2. Add a context layer from tables

    Pick tables from your warehouse and describe them so the agent gets accuracy on the metrics that matter.

More tutorials

The Bruin Cloud module covers each feature in its own short tutorial - configuring agents, scheduling runs, MCP, and more.