{"id":4789,"date":"2025-05-27T09:54:32","date_gmt":"2025-05-27T16:54:32","guid":{"rendered":"https:\/\/devblogs.microsoft.com\/semantic-kernel\/?p=4789"},"modified":"2025-05-27T09:57:27","modified_gmt":"2025-05-27T16:57:27","slug":"semantic-kernel-multi-agent-orchestration","status":"publish","type":"post","link":"https:\/\/devblogs.microsoft.com\/agent-framework\/semantic-kernel-multi-agent-orchestration\/","title":{"rendered":"Semantic Kernel: Multi-agent Orchestration"},"content":{"rendered":"

The field of AI is rapidly evolving, and the need for more sophisticated, collaborative, and flexible agent-based systems is growing. With this in mind, Semantic Kernel introduces a new multi-agent orchestration framework that enables developers to build, manage, and scale complex agent workflows with ease. This post explores the new orchestration patterns, their capabilities, and how you can leverage them in your own projects.<\/p>\n

<\/i>Optional reading<\/strong><\/p>Semantic Kernel Agents are now Generally Available<\/a><\/div><\/div>\n

Why Multi-agent Orchestration?<\/h3>\n

Traditional single-agent systems are limited in their ability to handle complex, multi-faceted tasks. By orchestrating multiple agents, each with specialized skills or roles, we can create systems that are more robust, adaptive, and capable of solving real-world problems collaboratively. Multi-agent orchestration in Semantic Kernel provides a flexible foundation for building such systems, supporting a variety of coordination patterns.<\/p>\n

Orchestration Patterns<\/h3>\n
Semantic Kernel now supports several orchestration patterns, each designed for different collaboration scenarios. These patterns are available as part of the framework and can be easily extended or customized.<\/div>\n

 <\/p>\n

1. Sequential Orchestration<\/h4>\n

In the sequential pattern, agents are organized in a pipeline. Each agent processes the task in turn, passing its output to the next agent in the sequence. This is ideal for workflows where each step builds upon the previous one, such as document review, data processing pipelines, or multi-stage reasoning.<\/p>\n

Use case example<\/h5>\n

A document passes through a summarization agent, then a translation agent, and finally a quality assurance agent, each building on the previous output:<\/p>\n

\"sequential<\/a><\/p>\n

2. Concurrent Orchestration<\/h4>\n

The concurrent pattern enables multiple agents to work on the same task in parallel. Each agent processes the input independently, and their results are collected and aggregated. This approach is well-suited for scenarios where diverse perspectives or solutions are valuable, such as brainstorming, ensemble reasoning, or voting systems.<\/p>\n

Use case example<\/h5>\n

Multiple agents generate different solutions to a problem, and their responses are collected for further analysis or selection:<\/p>\n

\"concurrent<\/a><\/p>\n

3. Group Chat Orchestration<\/h4>\n

Group chat orchestration models a collaborative conversation among agents, optionally including a human participant. A group chat manager coordinates the flow, determining which agent should respond next and when to request human input. This pattern is powerful for simulating meetings, debates, or collaborative problem-solving sessions.<\/p>\n

Use case example<\/h5>\n

Agents representing different departments discuss a business proposal, with a manager agent moderating the conversation and involving a human when needed:<\/p>\n

\"group<\/a><\/p>\n

4. Handoff Orchestration<\/h4>\n

Handoff orchestration allows agents to transfer control to one another based on the context or user request. Each agent can “handoff” the conversation to another agent with the appropriate expertise, ensuring that the right agent handles each part of the task. This is particularly useful in customer support, expert systems, or any scenario requiring dynamic delegation.<\/p>\n

Use case example<\/h5>\n

A customer support agent handles a general inquiry, then hands off to a technical expert agent for troubleshooting, or to a billing agent if needed:<\/p>\n

\"Handoff<\/a><\/p>\n

5. Magentic Orchestration<\/h4>\n

Magentic orchestration is designed based on the MagenticOne<\/a> pattern invented by AutoGen. It is a flexible, general-purpose multi-agent pattern designed for complex, open-ended tasks that require dynamic collaboration. In this pattern, a dedicated Magentic manager coordinates a team of specialized agents, selecting which agent should act next based on the evolving context, task progress, and agent capabilities.<\/p>\n

The Magentic manager maintains a shared context, tracks progress, and adapts the workflow in real time. This enables the system to break down complex problems, delegate subtasks, and iteratively refine solutions through agent collaboration. The orchestration is especially well-suited for scenarios where the solution path is not known in advance and may require multiple rounds of reasoning, research, and computation.<\/p>\n

Use case example<\/h5>\n

A user requests a comprehensive report comparing the energy efficiency and CO\u2082 emissions of different machine learning models. The Magentic manager first assigns a research agent to gather relevant data, then delegates analysis and computation to a coder agent. The manager coordinates multiple rounds of research and computation, aggregates the findings, and produces a detailed, structured report as the final output.<\/p>\n

\"magentic<\/a><\/p>\n

*Image by AutoGen<\/a><\/em><\/p>\n

<\/i>More orchestration patterns<\/strong><\/p>Semantic Kernel is also actively exploring additional popular agentic patterns to further enrich the framework. We encourage the community to share feedback and propose unique use cases, helping us continuously expand and improve the orchestration capabilities. Your insights and experiences are invaluable as we work together to make multi-agent development even more powerful and accessible.<\/div><\/p>\n

Simplicity and Developer-friendly<\/h3>\n

All orchestration patterns share a unified interface for construction and invocation. No matter which orchestration you choose, you:<\/p>\n