Let's be honest: most of us are using AI like a glorified StackOverflow. We copy a messy JSON payload, paste it into ChatGPT, ask for a Python schema, and then spend ten minutes fixing the hallucinated data types. If your AI workflow consists of juggling three different browser tabs and manually reviewing basic code, you’re missing the actual leverage of modern AI.
AI isn't going to take your job, but the data engineer who figures out how to automate their own grunt work is.
In this talk, we’ll move past simple prompt engineering and explore how to use the latest agentic frameworks to dramatically accelerate your daily data engineering workflows—from data exploration to code review loops. We’ll skip the marketing slides and dive straight into a live-coding demonstration of the cutting-edge open-source and data ecosystem tools you should be using right now:

Omnigent: The brand-new open-source "meta-harness" by Databricks. We’ll show how to orchestrate multiple tools (like Claude Code and Codex) under one roof to handle complex engineering tasks—like having one model plan a pipeline migration, another implement it in a sandbox, and a third cross-review the diff.
Databricks Genie: How to spin up AI data assistants to automate data profiling, documentation, and ad-hoc analysis, effectively cutting down your "can you run this query for me?" ticket queue to zero.
Whether you’re skeptical of the AI hype or just haven't looked at the ecosystem lately, come learn how to build an autonomous engineering partner that handles the tedious tasks so you can focus on system architecture.