{"id":7532,"date":"2026-04-10T08:58:12","date_gmt":"2026-04-10T15:58:12","guid":{"rendered":"https:\/\/www.microsoft.com\/en-us\/microsoft-copilot\/blog\/?post_type=copilot&p=7532"},"modified":"2026-04-10T12:30:28","modified_gmt":"2026-04-10T19:30:28","slug":"automate-business-processes-with-agents-plus-workflows-in-microsoft-copilot-studio","status":"publish","type":"copilot","link":"https:\/\/www.microsoft.com\/en-us\/microsoft-copilot\/blog\/copilot-studio\/automate-business-processes-with-agents-plus-workflows-in-microsoft-copilot-studio\/","title":{"rendered":"Automate business processes with agents plus workflows in Microsoft Copilot Studio"},"content":{"rendered":"\n

Today we are introducing new capabilities in Microsoft Copilot Studio<\/a> that help you automate your business processes by mixing AI agents and workflows. Agents and workflows already exist in Copilot Studio as two complementary capabilities with unique strengths. Agents bring reasoning and adaptability; workflows bring structure and consistency.<\/p>\n\n\n\n

\n
Build agents and workflows in Copilot Studio<\/a><\/div>\n<\/div>\n\n\n\n

So how do you know when to use agents vs. workflows?<\/p>\n\n\n\n

It\u2019s no longer an either-or decision. Here\u2019s how to use agents and workflows together to combine strengths and reduce risks.<\/p>\n\n\n\n

What are agents and workflows?<\/h2>\n\n\n\n

Agents<\/strong> are flexible AI solutions<\/a> that rely on foundational models to act, share knowledge, and handle tasks. They are powerful precisely because they are flexible. They can interpret unstructured inputs, reason over context, and make decisions beyond fixed logic.<\/p>\n\n\n\n

However, organizations often need to know that repetitive parts of their processes will behave consistently, every time they run. Pure agent autonomy doesn\u2019t always hold up to that requirement in production.<\/p>\n\n\n

\"Screenshot<\/figure>\n\n\n\n

Workflows<\/strong>, by contrast, are powerful automations<\/a> that drive process execution with consistency and speed. They\u2019re designed to deliver the reliability that many business processes require.<\/p>\n\n\n\n

At the same time, rigid, rules-based automation has its own ceiling. It\u2019s nearly impossible to anticipate every potential input format, edge case, and decision-making context when building a workflow ruleset. Thus, when the workflow automation encounters something unexpected, it can\u2019t move forward.<\/p>\n\n\n\n

Two patterns for scaling automation with AI<\/h2>\n\n\n\n

While both agents and workflows have their strengths, we\u2019re seeing customers get the most value in Copilot Studio by combining the two. In practice, we\u2019re observing two patterns emerge in how customers apply Copilot Studio, and we’re continuing to deliver product improvements to strengthen and support them.<\/p>\n\n\n\n

Workflows that use agents<\/h3>\n\n\n\n

The first pattern is workflows that call agents<\/strong>. In these instances, the workflow provides the structure<\/em> for the business process\u2014the defined steps, branching logic, handoffs, and an audit trail. Meanwhile, the agent handles the parts of the process that require judgement<\/em>. This might include interpreting a document, synthesizing information from multiple sources, or deciding how to route an exception.<\/p>\n\n\n\n

Once the agent completes its work, control returns to the workflow, and execution continues predictably.<\/p>\n\n\n\n

To make it easier to add agents to workflows in Copilot Studio, we\u2019re introducing agent nodes<\/strong><\/a>: the ability for workflows in Copilot Studio to call an agent directly within a workflow.<\/strong> You can build a deterministic, reliable automation, and at the exact moment you need AI reasoning, the flow simply hands it off to an agent.<\/p>\n\n\n\n

Setting up an agent node inside a workflow is simple:<\/p>\n\n\n\n

    \n
  1. Create a workflow step called \u201cAdd an agent.\u201d<\/li>\n\n\n\n
  2. Select any Copilot Studio agent you’d like to be include in the workflow.<\/li>\n\n\n\n
  3. Provide the instructions or task the agent needs to fulfill, and include an option to contact a designated person<\/a> if specific clarification is needed.<\/li>\n\n\n\n
  4. Add the rest of the workflow\u2019s steps.<\/li>\n<\/ol>\n\n\n\n

    When you run the workflow, the agent will do its job at the appropriate stage, and then the rest of the workflow will automatically continue.<\/p>\n\n\n\n

    \"Screenshot
    Adding an agent node inside a workflow<\/figcaption><\/figure>\n\n\n\n

    When to use agents inside workflows<\/h4>\n\n\n\n

    Using agent nodes to include agents in your workflows unlocks scenarios that rigid automation alone can\u2019t handle. Some potential uses include the following:<\/p>\n\n\n\n