Arun Agarwal

Build a complete AI retail app entirely in Snowflake. Start by using Snowflake CoCo to generate SQL and dbt models for metrics such as CLV. Next, train a GPU-accelerated XGBoost model in Snowflake Notebooks, register it, and deploy it as a service. To handle OLTP, use Snowflake Postgres and sync data to Apache Iceberg on S3, demonstrating lakehouse interoperability with pg_lake. Finally, unify the stack with Snowflake CoWork, a conversational interface orchestrating Snowflake Cortex Analyst, Snowflake Cortex Search and your ML models for a seamless analytical experience.

Learn how to bring Snowflake CoCo and agents together to build intelligent workflows that combine structured and unstructured data. In this hands-on lab, you’ll create agents that help a sales analyst answer product-related questions by leveraging both types of data retrieval. You’ll use Snowflake CoCo alongside MCP Servers to build the tools these agents rely on, implement agents using best practices, and set up evaluation systems to measure their performance. We’ll also explore how to build these agents programmatically, and how to use evaluations and traces to assess and improve their effectiveness. By the end, you’ll be familiar with best practices for building agents with Snowflake CoCo and MCP Servers, and how to evaluate their performance with confidence.

Snowflake CoWork has evolved from conversational Q&A to an end-to-end insights platform. This session unveils capabilities reshaping how enterprises consume data. See how we move beyond "what happened" to answer complex "why" questions with root cause analysis for actionable insights. Explore our new visualization experience: charts you can save, subscribe to and share, replacing static dashboards with a seamless conversational workflow. We'll demo high-IQ features including zero-shot forecasting, anomaly detection with proactive alerts and memory for personalized answers. Whether you're a business user or data team, discover how Snowflake CoWork is becoming the hub for enterprise intelligence.

This hands-on bootcamp teaches participants to design and deploy intelligent data agents using Snowflake Cortex. These agents orchestrate multi-step operations by retrieving and reasoning over structured and unstructured data with Cortex Analyst and Cortex Search. The session covers securely integrating AI into enterprise workflows, managing governance, and driving faster, informed decisions with real-time insights. Attendees will build a functional data agent capable of processing natural language requests, retrieving enterprise data, and generating intelligent responses within Snowflake. Completion of the lab earns a custom bootcamp badge.