Équipe Autogérée

The SelfManaged Crew component represents a collaborative group of AI agents, designed to work together to achieve complex goals by dividing and managing tasks. This component is ideal when you want to create workflows where multiple agents, each with distinct capabilities, can coordinate, execute, and manage hierarchical or multi-step tasks autonomously.

What the Component Does

• Orchestrates a crew of agents: You can define multiple agents, each potentially with different specializations or roles, to work together as a team.
• Task assignment and management: Tasks can be structured hierarchically, allowing for complex workflows to be broken down and assigned to the appropriate agents.
• Manager agent support: Optionally assign a specific agent to act as a manager, overseeing the crew and ensuring tasks are allocated and completed effectively.
• Customizable LLM for management: Specify a language model (LLM) for the manager agent, allowing for advanced coordination, reasoning, and communication within the crew.

Inputs

The SelfManaged Crew component offers a range of configurable inputs to tailor the teamwork and task management to your needs:

Additional Input Details

• Show Progress: When set to true, you can monitor exactly what each agent is doing at every step, which is particularly useful for debugging or process transparency.
• Max RPM (Requests per Minute): Useful for rate-limiting when interacting with APIs or external services.

Output

• Output: The component outputs a Message object, typically containing the results of the coordinated crew task execution. This can include the completion status, data generated, or final reports compiled by the crew.

Use Cases & Benefits

• Complex Task Automation: Perfect for scenarios where a single agent is not sufficient, and multiple agents with varying skills need to collaborate.
• Hierarchical Workflows: Useful for breaking down large problems into sub-tasks, assigning them to specialists, and aggregating results.
• Human-Like Project Management: Mimics real-world team dynamics, with a “manager” agent overseeing the process, providing oversight, and making decisions.
• Process Transparency: Enable progress tracking to see the inner workings of your AI crew, helpful for both development and production monitoring.
• Efficiency: Caching and rate-limiting options help optimize performance and control resource usage.

When to Use

Consider using the SelfManaged Crew component when your AI workflow requires:

• Multiple agents collaborating on interdependent tasks.
• Task breakdown and delegation mimicking human project management.
• Clear tracking and reporting of progress and results.

Further Documentation

For more detailed examples and advanced setups, refer to the official documentation.