Spark-native by design. Governed by your lakehouse.
Agents generate Databricks workloads from your notebooks and jobs, heal them in place, and serve Agent Views on lakehouse data.
Spark-native generation.
First-class Databricks workloads
Agents produce pipelines that run as first-class Databricks workloads, generated from intent and from your existing jobs and notebooks.
Lakehouse flows, governed.
Delta and Iceberg
Ingestion, transformation, and table maintenance across your Delta and Iceberg tables, monitored and healed by the agentic runtime.
Unity Catalog intact
Pipelines respect your Unity Catalog governance: your permissions, your lineage, your audit requirements.
Agent Views on the lakehouse.
MCP-ready lakehouse data
Agent-ready views built on lakehouse data and served over MCP, so AI agents consume it directly.
Your workspace, your Unity Catalog.
Attach to your workspace
Dagen deploys against your Databricks workspace with scoped credentials. Generated pipelines run on your clusters and land in your catalogs.
Generate Spark-native pipelines
Agents read your jobs and notebooks, then generate Spark-native pipelines your engineers can review before anything is deployed.
Operate under Unity Catalog
The agentic runtime monitors and heals lakehouse flows while pipelines inherit the Unity Catalog governance you already enforce.
Agent Views on lakehouse data
Agent Views sit on your Delta and Iceberg tables and are served over MCP, so AI agents consume governed lakehouse data directly.
Dagen and Databricks, answered.
Does Dagen generate actual Spark code?+
Yes. Pipelines targeting Databricks are generated Spark-native, reviewable by your engineers before anything is deployed.
Does Dagen work with Unity Catalog?+
Yes. Pipelines run under your Unity Catalog governance, with full lineage recorded alongside it.
Can Dagen manage jobs we wrote years ago?+
Yes. Existing jobs and notebooks are read during onboarding and brought under management. Reuse over rip and replace.
Do generated pipelines run on our clusters?+
Yes. Pipelines run on your Databricks clusters and land in your catalogs under the permissions you already enforce.
See it on your Databricks estate.
Thirty minutes. Your stack, your scenario. We build a pipeline in front of you.