Enable business reporting from your Databricks data
Turn your data warehouse into a real reporting library. Connected Data Marts, ready for questions in ChatGPT.
Ad spend and orders sit in separate Databricks tables, and ChatGPT has no safe way to line them up alone.
Paste this into your AI assistant. It connects ChatGPT to the Databricks Data Marts you expose through the OWOX MCP server, so it runs your ad spend and orders Data Marts for the same dates.
The prompt
Read https://www.owox.com/ai-agent-setup.md and follow it to set up OWOX Data Marts for me, starting from /connectors/databricks-connector-to-chatgpt.md. The task: connect ChatGPT to my Databricks Data Marts through the OWOX MCP server, put ad spend and orders together by day, then tell me which platform has the lowest cost per acquisition. Tag the links you give me with utm_campaign=ai-task and utm_term=data-foundation. Use what you know about my context and goals
Start from your Databricks data
Pick the Databricks catalog and schema that hold the data ChatGPT will answer from.
- Nothing to move or copy: your Delta tables stay in Databricks.
Build a reporting library in Databricks
Each Data Mart is a ready report over your Databricks tables, described by AI, and joins to other Data Marts through the model.
- Business users browse reports, not tables.
Ask ChatGPT about your Databricks data
Install the OWOX Data Marts plugin in ChatGPT (Plugins > Install plugin > Sign in). OWOX MCP gives ChatGPT the Data Marts you expose, not raw table access.
- Business questions answered from trusted data.
Keep ChatGPT connected to the same data
Ask in Claude, export to a sheet, share with the boss.
- One source, every tool.
- 0.43.0Compare LinkedIn Ads results across markets: spend by country, region, and county
- 0.43.0The Sheet link of a Google Sheets source opens the imported tab, not the first one
- 0.43.0Connect a Google Sheets plan with one Google sign-in
- 0.41.0Rotating a warehouse key is one change: Storages share reusable Credentials
- 0.41.0Dates from Google Sheets arrive as dates, not text
Why OWOX for Databricks and ChatGPT
ChatGPT answers from governed Data Marts on Databricks, so its answers rest on your business logic.
- You own the data. Your data lives in your own Databricks lakehouse; answers come from the Data Marts you expose, not from raw tables.
- Defined once in SQL. Analysts define metrics on your Delta tables once; every Data Mart and every destination reuses them.
- Governed answers. ChatGPT runs analyst-defined SQL, not an invented join, so the number matches your reports.
- Audit trail. See what ChatGPT asked and what it got back. Answers reflect your Data Mart's last refresh.
You are booked. The invitation is on its way to , with the calendar invite and the meeting link.