Daniel Pittner

Today @Motherduck , the company behind @duckdb, announced that they've raised $100mm. Yesterday, @tabulario, the team behind Apache Iceberg, announced a fresh $26mm round. And last week Databricks added $500mm to its coffers. What is happening?! What I believe we are now witnessing is the commoditization & fragmentation of the enterprise data stack, driven by open-source tools. This will be just as disruptive as the commoditization of the compute stack driven by Linux and x86 servers a couple of decades ago. Enterprise-grade data stacks can now be assembled from interchangeable, license-free software parts. Data orchestration, storage, and query tools are commodities. In the wake of this trillion-dollar wave of disruption, investors see opportunity, principally in building fully-managed (often "serverless") cloud services on top of the leading commoditized pieces. But creating these services, building community, and establishing a standard is a highly-competitive, multi-year undertaking. Hence the huge amounts of capital we're seeing flow into competing pieces of the stack. [In the analytics database category alone, there are dozens of well-funded, open-source commercializers vying for various slices of the pie -- MotherDuck (DuckDB), Clickhouse, StarRocks, Imply (Druid), Velox, Materialize, InfluxDB, TimeScaleDB, Rockset, StarTree (Pinot)... with an all-Rust-based option surely on the way somewhere.] As this disruption wave plays out, one thing OSS companies must engage with are the developer communities, who are innovation-minded, mission-motivated, and allergic to vendor pitches. On the path to building their businesses, I'm hopeful that the ultimate winners will follow the advice of Silicon Valley sage Tim O'Reilly: "create more value than you destroy." Further Reading: * The Road to Composable Data Systems: Thoughts on the Last 15 Years and the Future by @wesmckinn wesmckinney.com/blog/looking… * The Case for Independent Storage by @TabularBlue tabular.io/blog/the-case-for… * Building a Serverless Query Engine from Spare Parts (VLDB paper) by @GreCo_CiRo arxiv.org/pdf/2308.05368.pdf