Today I was also reminded that there’s no “easy” substitute for relevance.
We did some math to work out what the cost to roll out vector search would be… we’d need to increase our compute budget nearly 50%…. Just to roll out semantic search for issues.
We did some math to work out what the cost to roll out vector search would be… we’d need to increase our compute budget nearly 50%…. Just to roll out semantic search for issues.
Comments
I’m fairly sure a majority of our (human generated) searches are less than five terms.
This is another challenge is we have a huge population of searches that use massive generated queries.
It’s a use case we want to support but it’s not something we want to “waste” vector search compute on.