Concepts

Ascend is the Agentic Data Engineering platform. It combines AI-powered development with pushdown execution to your data warehouse, enabling data teams to build, deploy, and operate data pipelines faster.

At the heart of Ascend is Otto, an AI agent that accelerates every stage of data engineering. Otto provides code completion as you write, explains errors in plain language, and answers questions about your data lineage. In agent mode, Otto works autonomously to complete complex tasks — and you can build custom agents for domain-specific automation or configure background agents to handle routine operations like monitoring and alerting.

An Ascend Instance is the top-level container for all your resources. Within an Instance, Environments act as security boundaries — typically Development, Staging, and Production — each with its own access controls and Vault for secrets. Data engineers work in Workspaces, where they can edit code and iterate on pipelines. When ready, they promote changes to Deployments, where code is read-only and Automations run your pipelines on schedules or in response to events.

The code you write lives in a Project, which defines your Flows, Connections, and Automations in a Git repository. A Flow is a data pipeline — a graph of Components that run together. Components are the building blocks: Read Components ingest data from external sources, Transforms process data using SQL or Python, and Write Components export results to external destinations. You can also add Tests to validate data quality and Tasks for custom operations.

All of this runs on your Data Plane — a data warehouse like BigQuery, Databricks, MotherDuck, or Snowflake where your data lives and compute happens. Ascend acts as an orchestration layer, pushing queries down to your Data Plane's native engine for maximum performance while tracking metadata, lineage, and pipeline state in the Instance Store.