Adding agents for orchestration

Agents in IBM watsonx Orchestrate can collaborate by using multi-agent orchestration to complete tasks across various channels. Add multiple collaborator agents from IBM, your team, third-party platforms, watsonx.ai, and IBM watsonx Assistant instances, including prebuilt and custom agents. For more information on the list of available agents, see List of prebuilt agents.

Collaborator agents are computational systems where multiple autonomous agents interact or work together to complete specific tasks or achieve common goals. These agents can learn reasoning and decision-making.

Adding collaborator agents helps you to have domain-focused agents that are designed to operate within a specific field or industry. These agents possess deep knowledge and expertise in their respective domains, which allows them to complete tasks more efficiently and accurately.

With collaborator agents, you can:

  • Divide tasks among multiple agents to complete complex tasks more efficiently than a single large system.
  • Scale by adding agents to handle more tasks or larger workloads.
  • Enable the system to handle a broader range of tasks and adapt to changing conditions by allowing agents to carry out the tasks independently.
  • Gather insights from employee or customer communication, so agents can build upon key ideas for better performance.

Planning agents for collaboration and orchestration

Agents with collaborator agents operate independently and decide based on local information and interactions without a central authority.

An agent can have collaborators, and those collaborators can also have their own collaborators. This collaboration creates a network of interconnected agents. The initial agent can call upon all these collaborator agents to delegate the task execution to one another.

Using collaborator agents effectively requires careful planning and implementation:

  • Define clear objectives: Clearly define what you want to achieve with collaborator agents. Whether it's improving customer service, optimizing supply chains or enhancing data analysis, having specific goals to guide the implementation process.

  • Design the system: Plan the multi-agent system, including the roles and capabilities of each agent.

  • Choose the right agents: Choose agents that are suited to the tasks at hand. Consider their capabilities, compatibility with existing systems, and how they can be integrated into your workflows.

  • Foster human-agent collaboration: Promote a culture where human employees and agents work together seamlessly.
  • Plan for scalability: Make sure that your multi-agent system can scale as your business grows, which include you to be able to add more agents or expand their capabilities without significant overhauls.
Important: The collaborator agents in watsonx Orchestrate run sequentially. Currently, watsonx Orchestrate does not support parallel execution of collaborator agents or tools, and there is no configuration available to enable parallel processing.

Adding a collaborator agent

  1. On the agent edit page, click Toolset > Add agent.
  2. Choose from where you want to add the agent:
    1. Add from the catalog.
    2. Add from local instance.
    3. Import from external sources.

Choosing the right method

Method When to use Sample use case
From catalog When you want ready-to-use or partner-provided agents Adding a prebuilt HR or IT support agent
From local instance When your team already created agents in the same watsonx Orchestrate instance Reusing an internal finance or compliance agent
From external sources When you want to connect third party, watsonx.ai, or Assistant agents Using an AI model that is deployed on another platform
Using ADK When your team develops external A2A agents that must work with native ones Adding a custom logic agent that is built by using Orchestrate ADK

Using ADK to add external agents

You can build and register external agents by using the Agent Development Kit (ADK). For more information about how to build, configure, and register agents with ADK, see ADK documentation.

Removing a collaborator agent

  1. In the Toolset section, go to the agent that you want to remove.
  2. Click the vertical ellipsis overflow_menu next to the selected knowledge source.
  3. Click Remove.

What to do next

  • Specify how the agent should respond, including actions, restrictions, and settings for document-based chat, see Configuring the agent's behavior.
  • Connect your agent to channels so users can interact with it after deployment, see Connecting to channels.
  • Deploy the agent so it becomes active and available for use across the configured channels, see Deploying the agent.