Tutorials

Walkthroughs of end-to-end workflows on the OSO platform, aimed at getting you productive with a specific capability rather than a single API call.

Read top to bottom they form one path — get data out, shape it, present it, share it — and then hand that whole path to an agent:

1. 🔎 Querying OSO — run Trino SQL from Python (pyoso) or your agent (execute_sql), and discover what's available.
2. 📥 Ingesting data — bring external data into your org via REST, CSV, or Google Sheets.
3. 🧱 Data Modeling — author, materialize, and verify reusable SQL and Python models.
4. 📓 Notebooking — build a marimo notebook and publish it as a shareable snapshot.
5. 🌐 Publishing & sharing — public links, the data marketplace, and org collaboration.
6. 🤖 Automate with agents — connect an agent over MCP and let it run this whole path for you, with a verify-in-loop contract.
7. 🧠 Deep Research — point an agent at your data to produce a grounded, cited report.

The last two are the payoff: everything above is the manual path a hosted agent can run for you.

Looking for query examples and code snippets? See the data science tutorials for hands-on examples with pyoso.