Replies: 1 comment 1 reply
-
I would start with autogen core and layout all the components of the system, the events you think they can process, and the events you think they can emit. Model the events in proto. Write the event handlers for each top level event. Downstream from your top level orchestration choice you may choose to invoke agents working at the agent chat api layer, including potentially a multipurpose system like Magnetic One. |
Beta Was this translation helpful? Give feedback.
1 reply
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
Assume a chatbot design where human user can send various tasks/queries. Depending on the type of the task you may want to route the task to one of the two different teams:
There should be one orchestrating agent that selects which team to dedicate the task to. Additionally, any agent from any team can handoff to human user and ask for additional inputs or human confirmation.
How do we best design this in Autogen?
Beta Was this translation helpful? Give feedback.
All reactions