Elasticsearch as Vector Memory for Cheshire Cat AI
/ 1 min read
My colleague dej611 created a PR for the open-source Cheshire Cat AI project to add Elasticsearch as a vector database memory option in addition to the default Qdrant implementation.
Fun:
I have 0 experience with python, so sorry about silly mistakes in this PR.
At Elastic every few months we do what we call “ON Week” or “spacetime”, a week dedicated to learning new things or picking up side projects. We try to cancel all meetings and fully focus (don’t worry, we are still on call and support our customers!). It’s always a great opportunity to contribute something to the wider ecosystem like dej611 did here (and learning Python along the way). Over the years I did the following during ON Weeks:
- Picked up RE.GL to build a WebGL powered scatterplot.
- Learned Python as well with a focus on Jupyter Notebooks and data visualization and created jupyter2kibana.
- Developed and maintained d3-milestones.
- Contributed an Elasticsearch data loader to Observable Framework.
- Started to look into Rust (had a much harder time than picking up some Python!).