GoKawiilData warehouse startup drops DuckDB after scaling problems
The company behind a cloud data warehouse rebuilt its backend around DuckDB five months ago to give customers dedicated instances and a cost-based optimizer. After three months of re-architecting, it says it has removed DuckDB from its stack because customer workloads grew larger and more complex than its single-node, pre-sized instances could handle efficiently.
GoKawiil's interpretation of the reporting above, not reported fact.
The company's account suggests single-node engines like DuckDB, while excellent for local analytics, strain when deployed as multi-tenant cloud infrastructure that must predict resource needs before a query runs. This could signal broader limits in using embedded analytical databases as the backbone of hosted, pay-as-you-go warehouse products at scale.
- DuckDB's single-node design made it hard to size cloud instances before queries arrived.
- Complex customer workloads forced the company to use oversized, costly instances.
- The company also cited DuckLake's lakehouse format as built for compatibility rather than scale.
Source: posthog.com, 2026-11-02
Published there as: “Moving our warehouse off DuckDB”
Read the original report → The summary and analysis above are GoKawiil's own, written from reporting by the source above. Facts and quotes belong to the original publisher.