Semantic Kernel - Combine AI services & build your own copilot!
No, nothing to do with Kernel Sanders Chicken!
Although, if I wasn’t *mostly* vegetarian, I would be popping down to the local KFC! but I won’t…
So, this article has nothing to do with KFC, but it did mention “Kernel” so…
Microsoft semantic kernel? what the heck is it?
Well, for one thing, you can build your own copilots and more so, you can select to only use your own data or data sources (be that databases, PDF’s or other sources). As a brief primer (I will be doing a full walkthrough video soon!):
Semantic Kernel: An open-source SDK that lets you easily combine AI services like OpenAI, Azure OpenAI, and Hugging Face with conventional programming languages like C# and Python.
AI orchestration: A layer that allows you to combine AI models and plugins together to create brand new experiences for users. Semantic Kernel provides connectors and skills to add memories and models to your app.
Copilot experiences: Semantic Kernel allows you to leverage the same AI orchestration patterns that power Microsoft 365 Copilot and Bing in your own apps. You can also check out Project Miyagi, a sample of the copilot stack in action.
Open-source project: Semantic Kernel is available on GitHub in .NET and Python flavors (Typescript and Java are coming soon). You can contribute to the project by engaging in discussions, reporting bugs, or sending PRs.
In the walkthrough video (coming soon! yes, really!), I will also walkthrough some basics of “text embedding” using vectors and also a look at vector databases and why they have been, and will become more so, crucial to a lot of things “AI” :)