An open lakehouse with Git-like branches for data.
Tables are Apache Iceberg on your object storage, catalogued by Project Nessie. Every run lands on an isolated branch and merges to main only when its contracts pass — like a pull request, for data.
Zero-copy branch runs
A run writes to its own Nessie branch — a metadata-only operation, no data copied. Pass → an atomic merge to main. Fail → the output stays quarantined on the branch. Production has zero blast radius.
MERGE INTO upserts
Real upserts by key with Iceberg MERGE INTO — no full-table rewrites, no append-and-dedupe hacks. ACID by construction, safe under concurrent runs.
Time-travel by snapshot
Read any snapshot, branch or tag. Reproduce yesterday's report, audit exactly what a run changed, or roll a table back — all standard Iceberg metadata operations.
Queryable from Trino instantly
The moment a branch merges, the table is queryable from the built-in Trino — and from Spark, DuckDB or any Iceberg client. Analytics Studio ships on paid tiers for in-app SQL and dashboards.
No lock-in, by design
Open Iceberg tables on your object storage: if you leave SpooqW, your data doesn't move and stays fully queryable. The format is the contract — 100% open standards.
Your storage or ours
Cloud tiers include managed object storage and catalog. On-prem points at your own S3-compatible store, your own Nessie and your own Spark — the platform is identical.