Earlier this week, Microsoft launched the Agent Framework to its open supply challenge Semantic Kernel, an SDK that integrates LLMs with C#, Python, and Java in order that builders can incorporate AI options into their apps.
With this new experimental framework, brokers are actually supported within the platform. In line with Microsoft, brokers are AI entities specializing in having the ability to full a particular activity with a particular set of instruments. A number of brokers can work collectively to finish a posh activity as effectively.
Inside Semantic Kernel, every agent might be related to totally different plug-ins and capabilities that may assist it execute its job.
The Agent Framework consists of two particular agent varieties in addition to abstractions that builders can construct on. The primary agent sort is a Chat Completion Agent and the second is the Open AI Assistant Agent.
There are additionally two modes for interacting with brokers: immediately invoking it or by chatting with it by way of AgentChat.
Upcoming Agent Framework options which are being labored on proper now embrace help for the Open AI Assistant v2, enabling serialization and restoring AgentChat, improved chat patterns, and extra.
The Semantic Kernel challenge itself was first previewed in April 2023 and since then it has grown to 20K stars on GitHub.
Examine different current Open-Supply Tasks of the Week:
