Context is the opposite half of the story, and the context panorama is truthfully a multitude. Distributors are working arduous to persuade you that solely their semantic modeling language can prevent, whereas it’s not completely clear whether or not these are needed and even adequate. Whether or not you retain your small business logic in a semantic layer like MetricFlow or Malloy, or simply in plain Markdown, the aim is identical: get that logic right into a kind an LLM can use. Context is nearly at all times created by hand, and like all hand-written documentation, it begins drifting the second it will get written down.
This highlights a chance, specifically that brokers are good at exactly the elements of context which are mechanical and dangerous at exactly the elements that aren’t. An agent can infer which tables be part of to which, what values a column tends to carry, what your gross sales areas are, and which tables individuals really question. What it could actually’t infer is the stuff that was by no means actually a knowledge query: the proper approach to calculate income, what counts as a “buyer,” when the fiscal yr begins. These aren’t details hiding within the warehouse ready to be discovered. They’re selections, typically enterprise ones, that an individual has to make. What an agent can do is flag the second considered one of them quietly stops being true.
Automated agent insights stay a fantasy
The flashier pitch, the place brokers floor insights you by no means requested for, is the one I’d wager on final. It sounds great to have hands-free analytics. An agent will maintain watch over your information, discover what issues, and drop a dashboard tailor-made to no matter is going on in the present day. However the bar is excessive for relevance and false positives could make human customers lose confidence.
