Overview
A workspace is the core organisational unit in IntraLLM AI. It defines the scope in which models, knowledge, tools, and configurations are created and managed.
What a Workspace Contains
Each workspace may include:
- Models and training configurations
- Knowledge bases and documents
- Prompts and templates
- Tools, pipelines, and functions
- Chats and execution history
Resources created in one workspace are isolated from others unless explicitly shared.
Workspace Scope
Workspaces determine:
- Which resources are available
- Who can access and modify them
- How configurations are applied during execution
Operations such as training, document ingestion, and pipeline execution run within the context of the active workspace.
Permissions & Access
Access to a workspace is controlled by user roles and permissions.
Depending on your role, you may be able to:
- View workspace resources
- Create or modify configurations
- Execute tools and pipelines
- Manage workspace members
Administrative permissions are required for workspace-level management.
Switching Workspaces
Users with access to multiple workspaces can switch between them.
Switching a workspace updates the available resources and configurations shown in the interface.
Data Isolation
- Data and configurations are isolated by workspace
- Actions in one workspace do not affect others
- This isolation supports multi-team and multi-project usage
When to Use Workspace Basics
Use this page to understand:
- How work is organised in IntraLLM AI
- Why certain resources may or may not be visible
- How permissions impact what you can do
For guidance on navigating the platform, refer to Platform Walkthrough.
For layout and UI reference, see Interface Overview.
For workspace-specific configuration, see Core AI Capabilities, including Models & Training, Knowledge, Prompts, Document Extraction, and RAG.