Two Months In: Confessions of a New Engineering Manager

It’s been almost two months since I officially transitioned from Tech Lead to Engineering Manager.

I was reading The Making of a Manager recently, and I love that Julie Zhuo started writing down her learnings right as she transitioned, rather than waiting until she was an “expert.” There’s a specific kind of naive wisdom you have at the beginning that evaporates once you actually know what you’re doing.

So, before I grow out of this phase and start sounding like a corporate handbook, I wanted to write down a few thoughts.

Honestly? I think I tricked the system. I became an EM at the absolute perfect time.

The “Build” is Cheap, but Ideas are Expensive

One of the biggest fears people have about moving to management is losing the ability to build things. Materializing ideas is the best part of the job, right?

But we are living in a weird, beautiful window of time where “building” is becoming incredibly cheap.

The unexpected perk of being an EM right now is that I still get to build—I just build differently. In fact, I’ve found some convenient “excuses” to keep my hands dirty without the headache of deadlines.

Since I don’t have to pick up critical-path tickets anymore, I can use my expertise plus AI agents (shout out to Claude Code) to tackle the problems we never had the budget to solve. You know, the “high value, low urgency” stuff. Or the “extremely experimental, might-be-a-waste-of-time” stuff.

I can’t in good conscience assign a risky, experimental task to a report who needs a clear win for their promo packet. But me? I can take that risk.

I tell the team I’m building “comprehensive production instrumentation for future model training” or a “data analysis portal.” And sure, that’s true. But really? I’m just finding a way to scratch the builder itch using the latest tools.

The best part: I get to write code, but I don’t have to be on-call for it. (Don’t tell my boss.)

The $10 Challenge

Another thing I’m finding surprisingly fun is influencing the team—not by mandates, but just by being a nerd about tools.

I was an early adopter of GitHub Copilot, and now I’m deep into agentic workflows. I realized I could just “nudge” the team by being enthusiastic.

I jokingly set a goal for everyone: “Please, go waste $10 worth of credits on Claude Code.”

I didn’t ask for a report. I didn’t ask for a feature. I just wanted them to see what I was seeing. A week later, adoption spiked. It wasn’t a management strategy; it was just me sharing a toy I liked. It turns out that when you genuinely believe in something, “management” is just sharing that excitement.

The Late Night Shift

Connecting the dots between projects is great, and seeing the team grow is rewarding, but the most surreal part is the actual workflow.

It’s getting late as I write this.

On my work laptop, a Claude Code session is churning away on a task I set up earlier. On my personal laptop, I’m testing the latest Codex-5.3 model to build a feature for an open-source project I contribute to.

I am writing this blog post, but the code is still being written. The features are still being built.

I’m technically a manager now, but I’ve never felt more productive as an engineer. What a time to be living in, what a time!