dbt Agents overview BetaEnterpriseEnterprise +
dbt Agents, available on dbt Enterprise-tier plans, are a suite of native AI agents that turn structured dbt context into auditable actions. These agents help you build, manage, and consume governed data at scale by bringing intelligence to every step of the analytics development lifecycle.
dbt Agents are currently in beta or coming soon. Contact your account manager for early access.
See available agents to find out what's available.
dbt Agents are built on top of dbt's structured context to provide accurate, auditable, and governed results:
- Semantic Layer — Metrics, dimensions, and business logic
- Metadata — Lineage, tests, documentation, and ownership
- Governance — Access policies, data quality rules, and contracts
Having dbt as the standard context layer for agentic analytics means that dbt Agents are built on top of this context to provide accurate results rather than hallucinated or inconsistent answers.
Key benefits
- Faster development — Engineers and analysts ship data products faster with AI assistance.
- Better decisions — Business users get accurate answers grounded in governed data.
- Auditability — Every agent action includes transparent SQL, lineage, and policies.
- Scalability — Routine tasks are automated so teams can focus on high-value work.
Available agents
dbt offers several specialized agents, each designed for specific workflows in the analytics lifecycle to help you scale your data teams across the dbt platform.
dbt Agents are currently in beta or coming soon. Contact your account manager for early access.
Analyst agent Beta
Use Copilot to analyze your data and get contextualized results in real time by asking natural language questions to the Insights Analyst agent.
Chat with your data, get accurate answers powered by the dbt Semantic Layer. That means consistent, explainable results with transparent SQL, lineage, and policies.
The Analyst agent is a beta feature. Enable beta features under Account settings > Personal profile > Experimental features. For more information, see Preview new dbt platform features.
Discovery agent Private beta
Find the right, approved dataset fast in Catalog. The Discovery agent surfaces definitions, freshness, tests, owners, and lineage right where you work.
To request access to the Discovery agent, contact your account manager.
Observability agent Coming soon
The Observability agent in the dbt platform's orchestrator helps you monitor jobs, pinpoint likely root causes, and cut resolution time. It's designed to reduce noise and cuts down on investigation and debugging time — no more digging through logs.
Developer agent Coming soon
The Developer agent helps you describe the data question or product you want; the agent writes or refactors models, validates with Fusion, and runs against your warehouse with full context.
It helps you understand model logic, predict downstream impact, flag duplicate logic, and validate changes before merge. It runs directly in VS Code or Studio IDE, powered by dbt's context, so every change can be shipped quickly and safely.
dbt MCP server
Build your own custom agents and copilots with the local or remote dbt MCP server. The Model Context Protocol (MCP) makes dbt's structured context available to any AI system.
Related docs
Was this page helpful?
This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.
