Earlier this year, Microsoft released Semantic Kernel, which is an open-source SDK that allows developers to mix large language models (LLMs) with conventional programming languages.
Now the company has unveiled its roadmap for the project up through this fall. “Our focus is on delivering key developments and enhancements across three pillars: open source and trustworthiness, reliability and performance, and integration of the latest AI innovations,” John Maeda, vice president of design and AI at Microsoft, wrote in a blog post.
First, it has announced it is adopting the OpenAI Plugin open standard, which will allow users to create plugins that work across OpenAI, Semantic Kernel, and Microsoft’s platforms.
The company also announced updates to planner, which is “a function that takes a user’s ask and returns back a plan on how to accomplish the request… For example, if you had task and calendar event plugins, planner could combine them to create workflows like “remind me to buy milk when I go to the store” or “remind me to call my mom tomorrow” without you explicitly having to write code for those scenarios,” the documentation states.
Microsoft will add cold storage to planner, as well as introduce dynamic planners that will be able to automatically discover plugins.
The final plan in the roadmap is integrating Semantic Kernel with vector databases including Pine Cone, Redis, Weaviate, and Chroma. It will also integrate with Azure Cognitive Search and Services. Microsoft will also be adding a document chunking service and improving the VS Code extension.
“The fall 2023 release plan for Semantic Kernel emphasizes open collaboration, performance improvements, and the integration of cutting-edge AI innovations. Together, we are building a robust and reliable AI development platform,” Maeda wrote.