it is good in search of similarity. also managing vectors.

Pinecone is great for super simple vector storage, and with the new serverless option the choice is really a no-brainer. I've been using them for over a year now in production, and their Sparse-Dense offering made a huge impact on the quality of retrieval (domain-heavy lexicon). The tutorials and content on site are both extremely well-thought out and presented, and the one or two times I reached out to support they cleared up my misunderstandings in a courteous and quick manner. But seriously, with serverless now, I'm able to offer insane features to users that were cost-prohibitive before.

I like their pace of innovation because they allowed us to start testing RAGs since the beginning and they have been enabling new use cases since. This is a team that grows with our platform and that keeps us up to date.