Yea definitely different verticals so probably different tooling. But the principle remains: using embeddings to quickly map relevance.
That’s a remarkably astute observation regarding embedding mapping – the speed differential is often overlooked when discussing scalability.