Most organizations are structured the way they are because they were built for a world where humans did all the work.
Roles, departments, reporting lines, handoffs.
That world is changing fast, and most org charts are not keeping up.
IMO the focus shouldn't be on how you can retrofit AI into your existing structure. It's whether your structure still makes sense.
A method to think about it differently: Stop thinking in roles. Start thinking in actions.
For every function in your business, break it down to the individual actions that actually need to happen.
What can be fully automated? What needs human review? What actually requires judgment?
An ecommerce brand was drowning in customer inquiries - 150 a week. One person owned it all.
We ran an audit: 125 were routine and automatable.
25 were refund requests - 20 could route through policy rules + human sign-off, 5 needed actual judgment.
Result: 25 hours a week of manual work cut to 4 hours of high-judgment work.
Same person, but now with 80% more capacity.
Their response wasn't "we need less CX."
It was: how can we now unlock higher-leverage output from this person?
They concluded that their main focus would now be on post-purchase initiatives and driving LTV.
Some people will transition into higher-leverage work. Some roles may no longer make sense in their current form. Ultimately that is a leadership decision.
But you can't build a perspective on that until you have a clear picture on which steps in your workflow truly require human touch.
The companies that can do this early will have a structural advantage.
The ones that don't will be trying to layer AI into systems that were built for a different era.