Estuary’s Post

Iceberg is quietly becoming the standard for how serious data teams structure their lakehouses. Last week’s Cloudflare announcement was another proof point -- offering managed Iceberg tables and a built-in catalog. When a company like Cloudflare (at the scale and technical depth they operate) bets on Iceberg, it’s worth paying attention. And they’re not alone. 1. Snowflake has Polaris. 2. Databricks backs Unity Catalog. 3. AWS supports it via Glue, Athena and S3 tables. 4. Dremio, Starburst, MotherDuck, the list just keeps growing. At Estuary, we’ve been watching this play out, and more importantly, building around it. We see Iceberg as the table format that finally gives companies flexibility without tradeoffs. Governance without vendor lock-in. Performance without being boxed into a single processing model. But what’s been missing is great real-time integration. We’ve been working hard to fill that gap. Our Iceberg connector now supports: - Glue, Unity, Polaris, and other REST catalogs - S3 Tables - Streaming or batch ingestion from hundreds of sources - Schema evolution that just works - Merge queries for use cases that need precision (think CDC-style upserts) - And soon, built-in table maintenance If your team is starting to treat Iceberg not just as a format, but as an operational data layer—it might be time to talk. This is where the modern data stack is going and we're building for it.