May 12, 2026
I was recently talking with the CEO of a startup that is building AI agents for complex, process-heavy work. In their case, it’s tax preparation – one of the most time-bound, deadline-driven functions inside CPA firms. During peak season, their agents were freeing up, on average, roughly four hours per day per tax preparer. Not at the margins, but at the center of the work. And yet, the most interesting part of the conversation wasn’t the performance of the technology, but rather the response from the organizations adopting it.
They weren’t struggling to implement the agents. They were struggling to absorb the time.
In many cases, that newly available capacity didn’t translate into a rethinking of the work itself. Instead, it triggered…
Get the Full Story →