> ## Documentation Index
> Fetch the complete documentation index at: https://docs.prisme.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Overview

> A playbook for adopting and governing AI at scale: train, identify champions, open with a strategy, organize squads, and deliver critical use cases in project mode.

This is a **playbook, not product reference**. It is written from your point of view: *you* decide the strategy, *we* recommend how to run it, and every recommendation links to the product rubric that makes it enforceable.

## The operating flow

A pragmatic order that takes you from a first assistant to AI adopted and governed across the organization.

<Steps>
  <Step title="Train at scale">
    Start with acculturation for everyone: what AI can and cannot do, and the rules (AI Act). People govern themselves better when they understand them. It is the cheapest, highest-leverage step, and it comes first.
  </Step>

  <Step title="Identify and train AI champions">
    Use [Insights](/products/ai-insights/adoption) to spot the most active and relevant builders. Internal committees select the **AI champions**, give them deeper training, and grant them the right to share. Everyone else keeps building for themselves.
  </Step>

  <Step title="Open: choose your strategy">
    Decide how widely to open creation: **everyone builds (self-care)**, or **only selected use cases (curated)**. Then give access accordingly. By default no one shares or publishes; sharing is granted, never self-assigned. See [Opening strategies](./opening-strategies).
  </Step>

  <Step title="Organize the squads">
    Stand up the teams that keep the platform alive: a **Platform** squad (run and releases), a **Use-cases** squad (onboard complex use cases), an **Animation** squad (drive adoption), and **Architecture** (standards and reuse). See [Who does what](./operating-model).
  </Step>

  <Step title="Deliver critical use cases in project mode">
    A small set of critical use cases gets the full engineering treatment: the [use-case lifecycle](./use-case-lifecycle) and the environments and Git model in [Deployment strategy](./deployment). This is where infrastructure and CI/CD come in, for those use cases only, not for every self-care agent.
  </Step>
</Steps>

## Before you open: one org or several?

The structure decision most teams hesitate on, and where groups fit, is covered in [One organization or several](./org-vs-groups).

## The one principle

Governance runs on levels, and **each level can only tighten the one above, never widen it**. On conflict, the tightest limit wins. Inside its own envelope, each level distributes freely. It is detailed in [Opening strategies](./opening-strategies) and in [Governe](/products/ai-governance/overview).

## Explore

<CardGroup cols={2}>
  <Card title="Opening strategies" icon="door-open" href="./opening-strategies">
    Open to everyone or curate, and how to open without chaos
  </Card>

  <Card title="Who does what" icon="users-gear" href="./operating-model">
    The squads, committees, and champions, and who holds which right
  </Card>

  <Card title="One organization or several?" icon="sitemap" href="./org-vs-groups">
    The structure decision, and where groups fit
  </Card>

  <Card title="Critical use cases in project mode" icon="diagram-project" href="./use-case-lifecycle">
    The lifecycle and the deployment model for the use cases that need it
  </Card>
</CardGroup>