Ditch the Pipeline: Postgres to Snowflake Analytics with pg_lake

Operational and analytical workloads have historically lived in separate worlds, stitched together by brittle ETL pipelines that add cost, latency, and maintenance burden. As organizations demand faster access to operational data for analytics and AI, those pipelines have become a bottleneck. This lab is for data engineers, database engineers, and platform architects who work with Postgres today and want a simpler, more native path to Snowflake analytics.

To bridge this divide, pg_lake enables Snowflake Postgres to write data directly to Apache Iceberg format in Snowflake-managed storage — no ETL tools, no external pipelines, no data movement infrastructure to maintain. Snowflake reads those Iceberg files natively through a catalog integration, making Postgres data immediately available for analytics and AI. Incremental sync extensions built into Postgres handle the continuous refresh automatically, so new operational data flows to Snowflake in near real-time using nothing but SQL.

In this hands-on lab, we will demonstrate a live end-to-end workflow showcasing how Snowflake Postgres and pg_lake eliminate the need for traditional data pipelines: -Setting up Snowflake Postgres as an operational database and enabling pg_lake -Configuring incremental sync so new data flows to Snowflake automatically in near real-time -Creating a catalog integration to expose Iceberg tables in Snowflake — no ETL required -Running analytics directly on live operational data in Snowflake -Querying the data using natural language through a Cortex Agent

Attendees will leave with a blueprint for a modern, pipeline-free data architecture that connects operational and analytical systems natively. You'll learn how pg_lake and Apache Iceberg eliminate the ETL layer between Postgres and Snowflake, enabling near real-time analytics without the cost and complexity of traditional data movement. Join us to see how to go from a live Postgres database to AI-powered analytics in Snowflake.