A real day in the life of a ML engineer.

What actually happens between 9am and 6pm?

Most people picture an ML engineer at FAANG spending their day building models.

Training runs. Clean code. Elegant pipelines.

The reality is messier. More interesting. And honestly? More human than that.

Here’s what a real day looks like for me.

I’m at my desk by 7:30am.

Not because I have to be. Because I like it. The office is quiet, the coffee line is 0, and I can actually think before the day takes over.

Also useful trick: that’s when the MTV timezone night-owls people are still awake. Catching them before they go to sleep means I don’t lose a full day waiting on a reply.

The first thing I do is go through emails.

Simple rule: if the action item takes less than 5 minutes, I just do it immediately. Review a doc, answer a question, approve something. Done.

If it’s bigger, it goes on my notepad. Physical notepad. I’m not precious about tools, I just need somewhere to park things so my brain doesn’t have to hold them.

Then I look at the list and pick the most important thing. And I just go do it.

No elaborate prioritization system. No warming up. Action is everything for me.

I move fast, and I expect the same from the people around me.

Need a PR reviewed? Ping me. I’ll do it now.

Need feedback on a design doc? Send it. I’ll read it today.

I don’t want to be the bottleneck for anyone. I assume people like moving as fast as I do, so I treat their requests with urgency. It creates a kind of unspoken contract: I move fast for you, you move fast for me.

The pace compounds.

The actual ML work looks like this: