Workflow agents overview
Optimizely Opal workflow agents let you automate and optimize complex business processes by orchestrating triggers, logic, and agents. Workflow agents streamline multi-step tasks, support autonomous agent behavior, and enable advanced logic such as looping and branching. Workflow agents scale with your organizational needs and adapt to evolving processes.
Workflow agents are in private GA and not available in all Opal instances.
Structure
The following three sections work together to create a complete automation:
- Triggers – Define the events or conditions that start the workflow agent.
- Logic – Configure advanced logic, including loops, branching, and custom code, using specific conditions.
- Agents – Add and arrange specialized agents to create a complete workflow.
Together, these sections determine when a workflow agent begins, how it progresses through decision points, and which agents complete the process.
Specialized agents run in single-turn mode when used in a workflow agent, regardless of how they are configured. Workflow agents do not support multi-turn conversation. See Specialized agents overview for information on interaction modes.
When to use workflow agents
Workflow agents are ideal for coordinating multiple steps and connecting different agents into a single, automated execution. Workflow agents are most useful when a process requires measurable improvements in efficiency and accuracy.
- Streamline multi-step processes – Automate complex workflows with multiple steps and hand-offs to reduce delays, increase efficiency, and improve data consistency. This reduces time-to-market for campaigns and operational costs.
- Combine specialized agents – Let multiple specialized agents work together, applying reasoning and intelligent hand-offs to automate difficult or repetitive tasks. This reduces manual effort, minimizes errors, and frees up your team for more strategic work.
Best practices for workflow agents
Observe the following best practices to keep workflow agents efficient and manageable:
- Design modular agents – Keep specialized agents small and focused. Each agent should handle a single task or decision point. This modular approach makes complex workflow agents easier to manage and debug.
- Use workflow logic for complex processes – Avoid adding too much logic inside a single agent. Instead, use workflow steps, such as conditional branches and loops, to manage complexity and clearly show how the workflow agent progresses.
- Monitor and debug executions – Review execution logs to track how each step and agent performs. Regular monitoring helps you identify and resolve unexpected behavior.
- Choose the right triggers – Select triggers that make sense for the context of your workflow agent. Selecting appropriate triggers ensures your workflow agent starts under the correct conditions.
Next steps
See the following articles to create and configure workflow agents:
Note
If you use Opti ID, administrators can turn off generative AI in the Opti ID Admin Center. See Turn generative AI off across Optimizely applicationsTurn generative AI off across Optimizely applications.