TDWI Digital Dialogue | Building Reliable Agents on the Data Lakehouse
July 16, 2026
Success with agentic AI will depend on much more than selecting the right language model.
Organizations are moving beyond generative AI assistants to explore intelligent agents that can reason, plan, retrieve information, interact with enterprise systems, and take action. Yet despite considerable excitement, relatively few organizations have successfully deployed agentic AI into production.
In a recent TDWI webinar, TDWI VP of research Fern Halper spoke with Saurabh Shukla, senior specialist solutions architect at Databricks, and Ethan Post, principal field data engineer for AI observability at Monte Carlo, about what it takes to build reliable AI agents that organizations can trust. The discussion explored why many organizations remain stuck in pilot mode, the architectural capabilities required for production AI, the growing importance of observability, and operational best practices for deploying trustworthy agentic systems.
Download this Digital Dialogue today for the highlights of this discussion!