Dev Velocity Lab
← Back to blog
navigating-unfamiliar-roles

How I Went From Imposter to Actually Competent in My First SRE Role

I moved to SRE with zero confidence. My mid-year review surprised me. Here's what actually shifted—and it wasn't what I expected.

Watch on YouTube
ShareX / TwitterLinkedIn

Six months ago, I moved from full-stack development to an SRE team at Fannie Mae. I had no idea what I was doing. My previous role was me, maybe one other person, and AI filling the gaps. I'd ask ChatGPT about the docs, have it review my code, run my tests. My old manager? He'd ghost for weeks, then suddenly jump on a call when something broke and we'd all panic-brainstorm together.

Then I moved to SRE.

My new manager is involved. Like, in-the-weeds involved. He's in our technical standup, he understands our resiliency patterns, he talks to the app teams about priorities. The first few months felt suffocating. Everyone's in everything. Meeting after meeting. It's a team sport, not solo flights.

I went into my mid-year review expecting a "you're doing okay, keep learning" conversation. I got something different.

You Can't Learn in a Vacuum

My manager told me something that stuck: I needed to learn the business context of the 400 applications we support, not just the technical implementation.

This sounds obvious. It's not. When you're new and drowning, you focus on "how do I deploy this?" not "why does this app matter to Fannie Mae?" But when you talk to stakeholders, directors, other managers—they don't speak infrastructure language. They speak impact. Revenue. Risk. If I don't know what their app does, I can't translate what I'm fixing into terms they care about.

So now I ask questions. What does this team actually depend on? Where does the money come from? What breaks if we're down for two hours?

It sounds soft. It's not. It's the difference between being a checkbox engineer and being someone people trust.

Teaching Forced Me to Actually Understand

When I started teaching a teammate about our GitLab pipelines, SNS topics, Lambda triggers, and how they connect—I realized I'd been nodding along without actually seeing the shape of the system.

Teaching made me grab the specific commands. Trace the actual flow. Answer "why does this step happen here?" instead of "this step happens here."

Every time I explain something, it solidifies what I thought I knew but didn't.

The Shift Wasn't About Confidence

A year ago, I was afraid to admit I didn't know anything. I thought that made me an imposter.

Now? I know I don't know everything. And that's fine.

What changed: I show up focused. I ask questions without the shame. I pair with my manager and my team instead of hiding. I make mistakes and I learn from them instead of catastrophizing. Every week I'm hands-on with deployments, troubleshooting, AWS services, validations.

The imposter feeling didn't vanish because I suddenly became perfect. It vanished because I stopped expecting to be.

What Actually Works

  • Be totally present at work. Not distracted. Absorbed in what's happening.
  • Learn from the people around you, not just docs and AI. AI is a tool, not a replacement for real context from your manager and team.
  • Ask business context questions, not just technical ones. Why does this matter?
  • Teach what you learn. It's the only way to know if you actually understand.
  • Keep a positive attitude. Not fake-positive. Genuinely curious about what comes next.

I'm not an imposter anymore. I'm just a junior engineer on a steep learning curve, getting better every week. And honestly? That's enough.