I remember reading a comment somewhere about someone using datahike as a vector store and doing very fast in memory similarity searches. Anyone know anything about that or where I might find more info on how that might work?
Woops it was datalevin, i misremembered the name. Sorry to bother.