Hello! I've been thinking a lot lately about something that nobody talks about in those "day in the life of a software engineer" posts.

Everyone always tells you about the schedule. 9am standup. 10am coding. 2pm meetings. That stuff is boring and you can find it anywhere.

What I find interesting (and what nobody actually writes about!) is that the environment you work in is quietly training a different brain in you. Like, three years at Google produces a really different engineer than three years at a 12-person startup, and it's not in the ways recruiters pretend.

Here's what I mean:

1. big tech mostly trains you to read other people's code

If you work at a FAANG-tier company, most of your "coding" is actually reading code that 200 other engineers also touch. Stack Overflow's 2024 Developer Survey (which is a great source for this kind of thing) shows that developers at large orgs report some of the highest rates of working with legacy systems.

This is great training for some things – defensive thinking, writing for an audience of strangers, operating inside constraints you didn't pick. It's pretty bad training for other things – shipping speed, the muscle of deciding what to build, full-stack instinct.

I think this is actually kind of obvious once you stop and look at it, but it's not the thing recruiters or popular career posts talk about.

2. startups train you to ship fast and also kind of break things

At a 12-person seed startup, your day looks completely different. The CEO forwards you an email at 11pm because something is broken. You haven't had coffee yet. You're already in the code (which is a Next.js monolith plus some Cloudflare workers somebody wrote during a hackathon).

You fix it in 38 minutes. You ship it without a review because there's nobody to review it. You ship four more things before lunch.

What this trains: ownership, taste, pattern-matching across the stack. What this atrophies: rigor, long-term thinking, writing tests for code that "obviously works."

3. mid-size companies are actually the most interesting?

Mid-size companies (Series C through pre-IPO, maybe 200 to 2,000 engineers) get treated as a compromise but I actually think they're the most underrated. You ship real features for real customers, but there's enough process that you also learn what process is for.

A day here: standup with 7 people you actually know. Two hours of focused work. A design review where a senior engineer pushes back on your migration plan and you realize, in real time, that she's right.

What this trains: judgment. Specifically when to add process and when to skip it. You see both ends.

so which one is "right"?

This is the boring conclusion but: I don't really think there's a right answer. I do think that the best careers I've seen aren't loyal to one environment. They cycle. Two years at a big company. Three at a startup. Four at a mid-size. Then back.

The worst careers stay in one place too long and confuse the environment's blind spots for their own personality.

That's all!

If you're prepping for interviews across these environments, mock interviews tailored to your target company type are one of the best ways to calibrate your prep — the rounds are genuinely different.