My experience with (2) larger companies, and confirmed from conversations with friends, is that while you tend to have more support, you are often placed on individual siloed teams where you stay for years. I think the key benefit of smaller companies is that you can work across so many different aspects of the application. Process also tends to be less burdensome so you are able to devote more of your time to the text editor writing business logic.

Still, my experience and my circle is limited so I suspect some might have had the opposite experience.

Someone told me they spend 2 years in a team responsible for some of the undo menu at a giant software company. Incredible siloing.

To be fair, adding apparently simple undo functionality to a complex transactional system can be fiendishly difficult.

I was interviewing at a megacorp a long while back, my interviewer was solely responsible for the pie charts in some old application. He mentioned in the interview that it was totally feasible to run your own startup on the side.

An acquaintance of mine worked at Google for one entire year on the functionality of one single button in GMail (I don't recall what it was specifically). He was very disappointed, left and launched a startup.

My first job out of school was writing an OS from scratch for a new device. This was at a FAANG. Couldn't be happier. It depends on the opportunity.

That sounds like quite literally hell on earth. Thanks but no thanks (never mind the legal pandoras box of running your startup at your day job.)

Engineering learnings at big cos vs startups are wildly different. If your goal is to learn what it takes to start your own company, you are much better off working at even a mid-sized startup, than a big tech co. There are a number of reasons for this, but a particularly consequential one is that you will learn to build/manage/deploy applications from 0, and if things go well, through a variety of scales thereafter. At a mid-size startup you'll have less of the "from 0" bit, but you are much more likely to learn to deal with the scaling issues that manifest as the software that was built in the early days starts degrade with growing load. You will also likely be using similar open source tools in both the early & middle startup stages.

Building features in a large, existing codebase, in the robust tooling context of a big company, is nothing like building services from the ground up.

Larger companies have great programs, but also have large company problems which in many ways often aren't IT problems.