This session explores how AI builders can build and evaluate data agents. Learn how to connect unstructured and structured data sources to an agentic system, and how to evaluate the agent quality.
We’ll walk through how to build a data agent on Snowflake that has access to query structured data with Cortex Analyst and unstructured data with Cortex Search. We’ll also walk through how to evaluate an agentic system using Snowflake AI Observability, powered by TruLens OSS.
By the end, attendees will be able to build and evaluate an agentic system that can answer questions using different data sources, enabling them to deploy a data agent they can trust to production.
We’ll walk through how to build a data agent on Snowflake that has access to query structured data with Cortex Analyst and unstructured data with Cortex Search. We’ll also walk through how to evaluate an agentic system using Snowflake AI Observability, powered by TruLens OSS.
By the end, attendees will be able to build and evaluate an agentic system that can answer questions using different data sources, enabling them to deploy a data agent they can trust to production.
Josh Reini
Snowflake
Josh is a developer advocate for Open Source & AI at Snowflake. He is also a core contributor to open-source TruLens, an open source project for tracking and evaluating LLM apps with nearly 60k monthly downloads.
Josh has delivered tech talks and workshops to thousands of developers at conferences including PyData, AIDevWorld, Ai.Engineer, and a wide variety of technology and AI meetups.
Josh has delivered tech talks and workshops to thousands of developers at conferences including PyData, AIDevWorld, Ai.Engineer, and a wide variety of technology and AI meetups.