Fuel Every Agent and App with Data That Keeps Pace
Snowflake, Tuesday, November 3rd, 2026
At Expedition, Snowflake announced data engineering, analytics and transactional features to give AI agents and apps fresh, trusted data.
Snowflake argues that scaling AI depends on pipelines, tables and transactional systems that deliver current, trusted data, requiring a solid foundation plus AI that handles routine engineering work.
Announcements at its Expedition event include general availability of Adaptive Compute, which automatically right-sizes compute for workloads including Spark.
Snowpark Connect for Apache Spark, also generally available, runs existing PySpark, Java and Scala code on Snowflake without external clusters and can access Iceberg tables managed by Horizon, AWS Glue and Unity Catalog.
A Next-Gen File Reader for Apache Spark entered private preview for reading messy JSON, CSV, XML and Parquet files.