How it started (2019): I joined Scentbird as a Frontend Engineer — React.js, Node.js SSR, performance work. But from day one I could see the pain in code delivery. The company was growing fast and preached DevOps culture, yet there was no one who truly owned CI, infrastructure administration, or deployment pipelines. Jenkins was the CI — brittle, hard to configure, and opaque. Deployments ran through AWS Elastic Beanstalk: slow, nearly black-box releases with equally slow rollbacks. Every deploy was a gamble.

The experiments: In my free time between frontend tasks, I picked up a few infrastructure problems on my own initiative — just to see if I could solve them. I spun up the company's first Kubernetes cluster, migrated several services onto it, and wired up GitLab CI for transparent, fast deployments — then demoed the whole thing to the team. I also cracked a problem that had been unsolved for years: proper CORS configuration on S3 so we could drop the extra proxy that sat between users and S3 just to handle CORS headers. On top of that, I audited running infrastructure, shut down unused servers, and cut costs.

The transition: The results spoke for themselves — both sides won. I got a massive scope of new, genuinely interesting work; the business got someone who was solving long-standing problems fast. Within 4 months I fully transitioned from Frontend to SRE — the first SRE position the company ever had. Over the next 3 years I grew from SRE to Lead SRE, built out the entire infrastructure platform, and eventually hired and mentored my replacement before leaving.

Why I left (2022): After growing from SRE to Lead SRE, I reached a ceiling in infrastructure administration. Scentbird didn't have room for me to design backend systems from scratch — so I left to do exactly that at Azuro.

Why I returned (2025): Azuro and I parted ways as our interests diverged. Scentbird welcomed me back — this time with a broader mandate and a stronger SRE culture to build.