Music Recommendation System Pt 1

I love listening to music, but it's hard to find new music that I like. Over the years I have picked up a few tricks here and there, like following DJs whose taste I vibe with or looking up which record label put out an album I like and seeing what other artists have released music on that label.

With the advent of vector search databases, I now have a new tool in my toolbelt. In this series of posts, I am going to build a music recommendations site built on vector embeddings of reviews of music albums.

In order to save myself some time, I am going to use StageHand, a web automation library, to visit the music review site and get the text of music reviews into my application.

Then, I will use an embedding model to create vectors out of the text of the music review, and then upsert those embeddings into my vector database.

With my embeddings in my vector database, I can then build a front end user interface that interacts with my embeddings to recommend me music.

Let's get started!