I work at a fortune 100 insurance company in a dedicated IT subdivision of the company.
Fairly quickly, one might find that all requests are repetitive. It can be quite productive, professionally, to just assume that whatever you're asked to do, will be asked for again, and to learn how to pre-empt those requests by writing the first solution directly as automation code.
It doesn't have to be rocket science. Just keeping all the interesting bits in variables is the first 50% of automation.
You can become known as a miracle worker fairly quickly by just keeping a "commonplace book" of old one-off scripts, quickly found, in a personal git repo. Keep the company secrets left out to protect your skin, but a manager is often very quick to notice the person who can almost instantly replicate a previous request.
Having those snippets available also cross-pollinates - how many times have you repurposed one script into another? Careful now, once in a great while those turn into money making products, especially when the unrelated ones are hooked together.
These days, my little snippets are largely the boring, gross parts that act as the heavy barriers to automation - learning the URL endpoints for automation, learning the automation types, just getting a session that can do work. From there, the base case often yields a full strategy with another 10% of the work - modern APIs are so good you can generally just guess what is required and half the time it works without reading the docs.
Try coding it up. Worst case, you learn, best case, you invent a company. Middle of the road, though, you end up being a few people's go-to person, a wonderful place to be in this increasingly competitive world.