Don't want to set this up manually? Paste a link to this page into a Devin session and ask it to set everything up for you. Dana needs access to at least one database to answer questions. Go to **Settings > [MCP Marketplace](/work-with-devin/mcp)** and enable a database MCP: | Data source | MCP name | What you'll need | | --------------- | ---------- | ----------------------------------------- | | Amazon Redshift | Redshift | Host, port, database, username, password | | Google BigQuery | BigQuery | OAuth (one-click) or service account JSON | | PostgreSQL | PostgreSQL | Connection string | | MySQL | MySQL | Connection string | | Snowflake | Snowflake | Account, username, password | Click **Enable** on the MCP card, fill in your credentials, then click **Test listing tools** to confirm the connection works. You can connect multiple databases — Dana will query whichever one is relevant to your question. [Dana](/work-with-devin/data-analyst) is Devin's data analyst mode — a specialized agent with tools and prompts optimized for querying databases, building charts, and answering business questions in plain English. Open any Slack channel where [Devin is installed](/integrations/slack) and type `/dana` followed by a plain-English question. If the `/dana` slash command isn't available in your workspace, you can also use `@Devin !dana ...` — both work the same way. Dana creates a session, writes the SQL, runs it, and replies in-thread. ```txt Ask Dana a question theme={null} /dana What were our top 10 customers by revenue last quarter? ``` You can also invoke Dana by mentioning Devin with the `!dana` macro: ```txt Use the !dana macro theme={null} @Devin !dana How many active users do we have per plan tier this month? ``` Dana responds directly in the Slack thread with formatted tables, counts, and — when useful — a short interpretation of the results. No context-switching to a BI tool required. Dana isn't limited to quick lookups — it's also great at comprehensive reports and deep analysis. Ask it to investigate trends, segment data, and synthesize findings across multiple queries: ```txt Deep analysis theme={null} /dana Conduct a deep investigation on our user growth over the past month. Which user segments and use cases can the growth be attributed to? Is it more growth in existing accounts (expansion) or top of funnel (new signups)? Break it down by acquisition channel and plan tier, and highlight any notable changes compared to the prior month. ``` Dana works out of the box, but it gets significantly better when it knows your schema and business definitions. Instead of manually adding Knowledge entries, just give Dana feedback directly — in Slack or the webapp — and ask it to remember things for next time: * *"Remember that 'active user' means at least one event in `analytics.events` in the past 30 days."* * *"All our timestamps are stored in UTC. Always convert to America/New\_York for display."* * *"`billing.subscriptions.mrr` is in cents, not dollars. Divide by 100 when showing revenue."* * *"Update your knowledge: the `plan` column values are free, pro, and enterprise — not basic/premium."* Dana can list and update its own Knowledge, so corrections compound over time. The more feedback you give — column meanings, enum values, timezone rules, business definitions — the fewer follow-up questions Dana needs to ask in future queries. Dana keeps context within a Slack thread, so follow-up questions can reference earlier results without repeating yourself. ```txt Follow up on results theme={null} Break that down by plan tier and show the month-over-month trend. ``` ```txt Request a visualization theme={null} Can you chart daily signups for the past 90 days with a 7-day moving average? ``` For questions you ask regularly, consider creating a [playbook](/product-guides/creating-playbooks) with the exact queries and formatting you want, then trigger it with a custom `!macro` from Slack.

{title}