Atlan products
Atlan, the context layer for AI, offers two products. The Agentic Data Catalog is Atlan's enterprise context layer for your data estate, and Atlan Trident is the AI registry for your agents and skills. Learn what each one does and where its documentation lives.
Atlan is the context layer for AI. It offers two products:
- Agentic Data Catalog is Atlan's enterprise context layer. It connects your data estate, captures your business knowledge, and serves governed context to people and AI agents. This site, docs.atlan.com, documents it.
- Atlan Trident is Atlan's AI registry. It keeps your organization's AI agents and the skills they use in one place, records what they did, and controls which models and tools they can reach. Its documentation lives at platform.atlan.com.
Your data has context, and so do the agents that use it. The Agentic Data Catalog holds the first. Atlan Trident holds the second.
Compare products
| Agentic Data Catalog | Atlan Trident | |
|---|---|---|
| Built for | Data, governance, and analytics teams that need trusted, AI-ready data | Platform, engineering, and AI teams that build, share, and run agents |
| Answers | What does this data mean, where did it come from, and who can use it? | Which agents and skills exist, what did they do, and what did it cost? |
| Core objects | Assets, lineage, glossary terms, tags, data products, and policies | Skills, agents, metrics, traces, evals, and releases |
| How AI uses it | Agents read governed metadata through Atlan MCP | Agents publish skills, send traces, and reach models and tools through AI Gateway |
| Talk to data | Build a semantic model from governed context and deploy it to Snowflake, Databricks, or dbt with Context Engineering Studio | Give analytics agents your metrics and trusted SQL in Metrics Studio, then measure and improve their accuracy with evals and traces |
| Documentation | docs.atlan.com | platform.atlan.com |
Both products help with talk-to-data agents. If you need a semantic model in your data platform, build it from governed context in the Agentic Data Catalog. If agent accuracy is the main goal, use Metrics Studio with evals and traces in Atlan Trident. You set up each one separately. To take an analytics agent from demo to production, follow the method in Atlan's open-source analytics agents playbook.
Agentic Data Catalog
The Agentic Data Catalog builds the context your AI needs about your data. It crawls metadata from 80+ sources into your Enterprise Data Graph, tracks lineage, and lets you govern access, quality, and meaning in one place. Context agents enrich that metadata at scale, and MCP serves it to the AI tools your teams already use.
To learn more, start with What is Atlan?, or go straight to Context Agents Studio, Context Engineering Studio, and Atlan MCP.
Atlan Trident
Atlan Trident is where your organization keeps its AI agents and the skills they use. It works beside the tools your teams already use, such as Claude Code, Codex, Cursor, agent frameworks, and your own code. With Atlan Trident, you can:
- Share skills. Publish versioned skills from a GitHub repository, the desktop app, or the command line, and see who owns each one and which version is live.
- Register agents. Record each agent with its versions, skills, sessions, and runs.
- Observe and evaluate. Collect traces, see usage, errors, and cost per agent, and test changes against fixed cases.
- Give analytics agents your metrics. Define your team's metrics and trusted SQL in Metrics Studio, so talk-to-data agents answer business questions with the numbers your team agrees on.
- Govern access. Route model calls and MCP tool calls through AI Gateway, and control which models and tools each agent can reach.
Atlan Trident was previously called AI Registry and Agent Registry.
Atlan Trident is in private preview. See Availability and limits for what's on by default.
To learn more, see What is Atlan Trident.
Open source from Atlan
Atlan also publishes two open-source projects:
- Registry OSS is an open-source agent registry: a specification that makes agent registries an open standard, and
aireg, a CLI that works with any registry that follows it. Both use the Apache 2.0 license. You can run a registry on storage you already have, such as a Git repository. Atlan Trident is Atlan's hosted product, which adds workspace permissions, SSO, approvals, and traces and evals. See the project documentation and the repository. - Analytics agents playbook explains how to build an analytics agent that gets your business numbers right: metrics from your analysts' SQL, a golden set of real questions, and improvement loops. Its text is free to reuse under CC BY 4.0. See the playbook and the repository.