Query your data lake with SQL — no warehouse.

Run SELECT queries over Parquet, CSV, JSON and Iceberg in object storage, on DuckDB or Trino, and pay only for the bytes you scan.

Submit a query, then fetch results

# submit an async query over Parquet in object storage
POST /api/query-executions
{ "sql": "SELECT count(*) FROM read_parquet('s3://bucket/events/*.parquet')",
  "workgroup": "primary", "engine": "duckdb" }

# poll status, then page the results
GET  /api/query-executions/{id}
GET  /api/query-executions/{id}/results?maxResults=1000