What is a Crew
A Crew is the top-level container for your AI agents and their configurations. Think of it as a project or workspace that holds everything related to a specific use case—agents, knowledge bases, pathways, and integrations.Anatomy of Crew
Every Crew contains the following components:Agents
Agents are the AI personas that interact with your users. Each agent has:- A persona (name, voice, personality)
- Behavior rules (guardrails, escalation triggers)
- Capabilities (call transfer, appointment booking, data collection)
Knowledge Base
The knowledge base contains information your agents can reference during conversations:- FAQs and standard responses
- Product information
- Company policies
- Structured data (pricing, hours, locations)
Pathways
Pathways define the conversation logic and decision trees your agents follow. They determine:- How conversations branch based on user input
- When to escalate to a human
- What actions to take (book appointment, send SMS, update CRM)
Multi-Crew Architecture
Enterprise deployments often use multiple Crews:| Use Case | Crew Structure |
|---|---|
| Single business | One Crew with multiple agents |
| Multi-location | One Crew per location |
| White-label platform | One Crew per client |
| Dev/Staging/Prod | Separate Crews per environment |
Crew vs. Workspace
A Workspace is the billing and access control container that holds one or more Crews. Users are invited to Workspaces, not individual Crews.
Creating Your First Crew
1
Navigate to Dashboard
Go to app.usecrew.ai and log in
2
Create Crew
Click New Crew and provide a name and description
3
Configure Basics
Set your timezone, default language, and business hours
4
Add an Agent
Create your first agent with a persona and knowledge base
Best Practices
- Separate environments — Use distinct Crews for development, staging, and production
- Descriptive naming — Name Crews by function or client for easy identification
- Limit scope — Keep each Crew focused on a specific use case or workflow
- Version control pathways — Export and version your pathway configurations
Next Steps
- Agents Overview — Learn about the different agent types
- Knowledge Base — Populate your agent with information
- Pathways — Design conversation flows