Semantic Kernel 2024 Spring Roadmap!

Поделиться
HTML-код
  • Опубликовано: 28 ноя 2024

Комментарии • 4

  • @masegraye
    @masegraye 9 месяцев назад

    Love the roadmap. I’ve been experimenting with my own implementation of agents that doesn’t take a dependency on the Assistant’s API. Goal-directed, planning, self-correcting, drawing on memory/experience (a la MemGPT), delegating to a panel-of-experts, etc.
    Some of my more fun experiments include creating an agent that knows how to create other agent definitions based on a high level task.
    Semantic Kernel has proven a useful foundational tool for building this. One thing I’m not a fan of is the hard dependency on the Azure SDK, which appears to have a fixed release cycle tied to .NET.
    To me, that’s untenable in the face of OpenAI’s rapid release cycle. If you want to own the space, you can’t take a dependency on something with a slower release cycle than the forcing function for innovation.

  • @shreyaspandey131
    @shreyaspandey131 9 месяцев назад

    Thanks for sharing the plan. Could the SK assistants be used with Azure Open AI as of today? If not, do we expect the support to be available soon?

  • @hacker0111-qw1vn
    @hacker0111-qw1vn 9 месяцев назад

    Love it!

  • @dipteshbosedb
    @dipteshbosedb 9 месяцев назад

    Will be great if the community is blessed with a Low Code way of using Semantic Kernel with a Semantic Kernel "Studio", like how Autogen team has provided with Autogen Studio.