oleander

The product is a multi-engine data warehouse with intelligent query routing that translates and distributes SQL or natural language queries to the most efficient compute engine to control cost and latency. It is for developers, data teams and IT teams in B2B companies building agent-driven and data-intensive applications, from small teams to enterprises. It is positioned as a fully managed, zero-ops platform that routes each query based on characteristics, context and cost history across engines such as DuckDB, Polars, Spark, Snowflake and Databricks.

Key features

• Intelligent query routing across engines
• Fully managed storage and compute
• Zero-ops warehouse launch in minutes
• Support for Bloom, DuckDB, Polars, Spark
• External Snowflake and Databricks support
• Managed Iceberg catalog
• Bring your own Glue and Iceberg catalogs
• OpenLineage-powered lineage tracking
• SQL telemetry and observability
• Lineage-powered governance
• Environment separation via routing
• User and agent level permissions
• Native agent plugins for Claude and Codex
• Universal MCP connection
• Agent guardrails for queries
• Automatic anomaly investigation and RCA
• Unified SDK for querying
• CLI for querying and management