A two-part series on building AI workflows that data engineers and analysts can actually trust. Use the agent for speed, use your judgment for accuracy.
Session Details
Part 1 · June 18
Foundational Principles and Data Engineering Workflows
Part 2 · June 25
Analytical Acceleration and AI-Powered SQL
Time
9:00 AM PT / 12:00 PM ET / 5:00 PM BST / 6:00 PM CEST
Duration
1 hour each · One registration covers both
One registration covers Part 1 and Part 2. We'll send the recording if you can't make it live.
Two sessions covering the full arc from foundational AI agent control to production SQL on unstructured data.
Describe, Plan, Execute, Verify isn't a slogan, it's a sequence you can use on every task. Part 1 walks through the loop applied to real data engineering work, then Part 2 stress-tests it on analytical edge cases. Leave with the loop committed to muscle memory.
Project files give the agent your conventions. MCP gives the agent live catalog metadata. Agent skills give it your business rules. Together, that layering stops hallucinations before they happen and gets you correct SQL dialects, naming conventions, and join semantics on the first try.
AI_CLASSIFY and AI_GENERATE let analysts process unstructured text, documents, and files inside SQL, no fragile Python pipelines to maintain. We'll show real query patterns and the verification habits that keep results trustable.
Three resources to extend what you build in the sessions.
Step-by-step guide to connecting AI agents to your Dremio Cloud project via the Model Context Protocol.
How the combination of MCP and a governed semantic layer turns agentic analytics from demo into something you'd put in front of customers.
Practical walk-through of AI_CLASSIFY, AI_GENERATE, and AI_COMPLETE, including the workflow patterns Part 2 builds on.