What changed
2026-08-18 12:07 UTC → 2026-08-27 05:31 UTC ·
added
removed
Snowflake expands AI‑ready AI-ready data ecosystem and agentic tools
Snowflake’s Open Warehouse architecture, demonstrated in July, links an Apache Iceberg lake on Amazon S3 to Snowflake compute via external volume definitions and AWS Glue catalog integration, enabling zero‑copy governance and semantic views while keeping data in S3. The move reflects a broader industry shift toward lake‑house designs that combine lake flexibility with warehouse performance. In the same period Snowflake released Data Metric Functions (DMFs), a native capability that continuously measures data‑quality indicators such as null counts, distinct values and table freshness. DMFs can be scheduled, store results in a dedicated event table, and support both system‑provided and custom metrics, reinforcing a unified semantic layer that delivers consistent definitions across dashboards, notebooks and AI interfaces. Snowflake‑led Open Semantic Interchange (OSI) gained a new partner, Solid, whose AI‑native context layer will automate creation and maintenance of semantic metadata, aiming to prevent semantic drift across AI agents and analytics tools. To broaden is advancing its AI Data Cloud capabilities, Snowflake announced collaborations with phData and Spotfire. phData will leverage Snowflake’s CoCo and Cortex platforms to accelerate enterprise AI production, aiming to reduce delivery times capabilities by up to 80%. Meanwhile, Spotfire’s Push Compute integration allows heavy-industry users to run large-scale transformations directly in Snowflake without moving data. Snowflake is also developing an internal semantic layer designed to act as an intermediary that translates between governed business language into and physical database schemas. This “golden layer” of data meaning aims to ensure that both humans and AI agents receive consistent answers to business questions. questions, such as defining an “active customer.” This development aligns with technical evolution reflects an industry trend of industry-wide shift toward rebranding AI offerings as “enterprise AI coworkers” that coworkers.” Rather than requiring users to visit standalone interfaces, these models integrate directly into existing workflows via connectors for platforms like Microsoft Teams Teams, Excel, and Excel. Google Sheets. This approach mirrors competitors like Databricks, which has introduced Genie One as an “agentic coworker.” At the Snowflake World Tour Seoul, executives emphasized that a robust data strategy is the essential foundation for the “agentic enterprise,” where AI agents perform business tasks by connecting scattered data across ERP and CRM systems under strict security controls. The event highlighted the practical scaling of these technologies; for example, KB Financial Group reported operating over 100 AI agents, with plans to expand to 300 this year. In South Korea, Snowflake has seen significant growth, with local revenue increasing approximately 18-fold since its 2021 launch.