GitLab is becoming an AI-first company. Enterprise AI is the AI backbone of that transition internally: the platform, the governance, and the operating patterns that let every team work with AI as a core part of their craft, not as a side experiment.
Concretely, the goal is friction removal. Team members should stop scavenger-hunting across tools for context, stop stitching threads together by hand, and stop repeating work because no shared instructions exist. Institutional memory becomes queryable. Repeatable work becomes an agent's job. Humans direct, agents handle the parts that do not need judgement.
What GitLab builds for customers in the agentic era, we live ourselves. Enterprise AI is GitLab being Customer Zero for the patterns we want our customers to feel.
## What we own
We own GitLab's AI platform (Glean, Claude, and Relevance.ai, listed below), the governance and standards that sit around it, and the cross-function patterns and enablement that turn tooling into outcomes. We are the single interface to CorpSec, Legal, and Privacy for AI tooling, so functions do not each negotiate those reviews on their own.
## How we operate
We work as a Hub and Spoke and Hub model. Enterprise AI is the platform Hub. Each function has a spoke led by an AI Transformation Owner (ATO), who connects platform capability to in-function priorities. Inside each function, a Hub of Champions carries practice and patterns into the day-to-day work. See [Hub and Spoke](/handbook/eta/ai/strategy/hub-and-spoke/) for the full operating model.
## Our tools
The foundational tool is GitLab itself. It is the platform we share with our customers, and we lead with GitLab and Duo Agent Platform wherever they fit. Plenty of what we build runs through the other platforms below as well; the right tool wins. We do not have a tool page for GitLab in this section because Enterprise AI does not own or operate it. That work sits with the engineering org, and with the [AI engineering team](/handbook/engineering/ai/) in particular. Where our work intersects with theirs, we partner directly.
Alongside GitLab, we operate three AI platforms ourselves, each for a different kind of work:
-**[Glean](/handbook/eta/ai/tools/glean/)**, GitLab's personal AI for knowledge management and personal-scope automation across our internal systems.
-**[Claude](/handbook/eta/ai/tools/claude/)**, GitLab's desktop AI for getting work done: drafting, reasoning, code, agentic tasks.
-**[Relevance.ai](/handbook/eta/ai/tools/relevance-ai/)**, GitLab's agentic AI platform for company-wide autonomous agents.
## How to engage us
- Join [**#enterprise-ai-collab**](https://gitlab.enterprise.slack.com/archives/C0AE5DX6SQJ) on Slack for questions, FAQs, and intake support.
- Submit ideas and use cases via the [Enterprise AI Intake Request](https://gitlab.com/groups/gitlab-com/eta/ent-ai/-/work_items/new?type=EPIC&initialCreationContext=list-route).
## Where we sit
Enterprise AI is part of Enterprise Technology and AI (ETA), under the CIO.
Enterprise AI is built on a clear operating model and a small number of durable principles. The pages below cover how the model works in practice and the reasoning behind it.
