AI Agents
rudol brings AI into your data platform in three ways: an agent you ask, a suite of agents you delegate work to, and an endpoint you connect your own agents to.
rudol's agents are grounded in your metadata and never change your catalog on their own. AI Crew proposes actions that a person approves, and agents do not generate a suggestion when there isn't enough evidence to support it.
Copilot
Copilot is the conversational agent built into rudol. Use it to ask about domains, glossary terms, metrics, lineage and any other catalog item in natural language — it answers using your catalog as context. You can personalize its tone, what it prioritizes and how it returns links to your team.
AI Crew
AI Crew is a suite of agents that do the item-by-item work of data governance for you: each agent analyzes the catalog according to a policy you write, suggests a concrete action, and a member of your team approves or rejects it. Use it to document or classify large parts of the catalog without reviewing every object by hand.
Connect your own agents (MCP)
rudol exposes a Model Context Protocol (MCP) endpoint so you can plug the catalog into your own AI agents and applications. Your agent can retrieve documentation, business context and other catalog knowledge from rudol, inheriting the permissions and catalog visibility of the rudol account it authenticates as.
Want to connect your own agent through MCP? Reach out at hello@rudol.ai and we'll walk you through setup.