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

# Organizations/projects

Nebuly uses a hierarchical structure to organize your AI analytics. Every workspace is composed of **Organizations** and **Projects**, allowing you to separate data, manage access, and analyze multiple AI applications from a single platform.

This structure is designed for organizations operating multiple AI assistants, copilots, or customer-facing AI products.

## Organizations

An **Organization** is the highest-level container in Nebuly.

Organizations provide complete isolation between environments, including data, users, authentication, and configuration. They are typically used to separate:

* Different companies
* Production and development environments
* Business entities requiring complete data isolation

Data never flows between Organizations.

## Projects

A **Project** is the primary unit of analysis inside an Organization.

Projects contain the interactions, conversations, users, reports, taxonomies, and analytics for a specific AI deployment.

Most organizations create one project per:

* AI assistant or agent
* Product
* Business unit
* Region or market

Projects are where you'll spend most of your time analyzing AI usage.

### Root projects

A **Root Project** stores the actual interaction data.

Every conversation ingested belongs to a single root project. Each root project has its own:

* Reports
* Taxonomies & ROI/Proficiency configuration
* Processing pipeline

Root projects are completely independent from one another.

### Sub-projects

A **Sub-project** is a filtered view of a root project.

Instead of storing duplicate data, a sub-project applies a permanent filter to its parent project while maintaining its own permissions.

For example, you could create:

* Customer Support → France
* Customer Support → Germany
* Customer Support → Italy

All three use the same underlying conversations but only expose the relevant subset.

Sub-projects are ideal for regional teams, departments, or business units that should only access their own data.

### Switching organizations & projects

Use the organization/project switcher to move between organizations and projects.

Nebuly remembers the last project you opened and restores it automatically the next time you sign in.

<Frame caption="Selecting a sub-project from the project switcher">
  <img className="rounded-xl" src="https://mintcdn.com/nebulyai/ahBVJk9XFVot7JV6/images/subproject-result.png?fit=max&auto=format&n=ahBVJk9XFVot7JV6&q=85&s=a9ea97232fe2fae8b01ab173f9c8eebc" alt="Selecting a sub-project from the project switcher" width="1112" height="776" data-path="images/subproject-result.png" />
</Frame>

### Comparing projects

Reports can compare two projects side by side, allowing you to analyze differences between AI deployments, teams, or regions without rebuilding your reports.Best practices

* Create one **root project** per AI application.
* Use **sub-projects** to segment the same data by region, department, or business unit.
* Use **Organizations** only when complete isolation of data, users, or configuration is required.

Read more about how to compare projects [here](/guides/reports/configure-reports).

## Managing organizations & project

To learn how to create projects, configure permissions, and manage organizations, see [**Managing Organizations & Projects**](/guides/setup/projects-and-organizations).