Presented by:Guy Royse
I’ve written a rather silly thing using AI and JavaScript that finds Bigfoot using Semantic Search. In this session I’ll show it to you, and show you how to build your own—all using local models, local databases, and local code—and without plopping down a credit card to use cloud APIs.
I’ll show you the code I used to trigger a Semantic Search. Not sure what Semantic Search is? Well, it uses embeddings to search for what you intended as opposed to keywords. This lets it understand the context of your request. Not sure what embeddings are? An embedding is a very carefully crafted vector created using an AI model. How can vectors do that? Guess you’re gonna have to come and find out!
If that’s not enough, I’m using buzzwords like: LLM, LangChain, Hugging Face, Redis, and Vector Database. Now you have to check it out.
So come for the buzzwords, stay for the fun, and leave with the knowledge you'll need to build something fun, useful, or both.
Level: Introductory and overviewTags:AI & ML, Databases