Using Quora questions to test semantic caching for LLMs

I've been using Elasticsearch to do a similar job, to mixed results.

What is it you're using to do the actual semantic indexing? I scanned the code but couldn't figure it out.