There's a way companies can save thousands, maybe even millions, on data reporting alone: paste the code behind your reports into an LLM and ask it to optimize it. It's one of my favorite LLM use cases.
Every report you look at each morning has several processes behind it, like transformations and aggregations, that get the data in front of you. Much of that is written in SQL, where there are a thousand ways to get the same output. Most people writing SQL aren't writing for efficiency, so the output is correct, but the code could easily run faster, which means cheaper data costs.
Do this and you'll save on data costs. Bonus: it's how I learned about the ILIKE function in Snowflake.