Before you start
Last updated
Last updated
Domain-Expert Agents (DXAs) are AI-powered assistants tailored to specific expertise domains. Before creating your agent, it's important to understand the different agent roles and capability packages available.
Agent Roles
Specialist Agent: Focused on a single domain of expertise.
Data Analyst Agent: Designed specifically to process and analyze numerical information from spreadsheets and data files.
Coordinator Agent: Functions as an orchestrator managing multiple specialist agents to tackle complex, multi-domain tasks. These agents break down problems, delegate subtasks to appropriate Specialist or Data Analyst agents, and synthesize comprehensive solutions.
Capability Packages
For each Specialist Agent, you can select from three capability levels:
Essential (default): The agent will conversationally answer questions using expert knowledge. Ideal for information retrieval and advisory roles where straightforward expertise is needed.
Operational: The agent will follow structured workflows to solve problems systematically. Best for complex tasks that benefit from methodical approaches and step-by-step solutions.
Autonomous (BETA featured): The agent self-directs complex problem-solving, making independent decisions about approach and execution. Suitable for advanced tasks such as research, diagnostics, and scenario analysis
Please refer to with different capability packages to learn more detail on how the agent performs with different capability packages.