Puneet Bharti

The Journey

I joined as a backend developer. At the time, I thought my world would mostly revolve around writing APIs, fixing bugs, and shipping features.

Then one night, production went down.

The whole dev team was trying to figure it out. Everyone was digging through the codebase, looking for answers. But something told me the problem wasn't in the application itself — it was deeper than that.

So I started exploring the systems underneath it all.

One SSH session at a time, I jumped across virtual machines, grepping through endless logs buried deep inside servers. No fancy observability stack. No dashboards. Just a black terminal screen, Vim, and patience. A lot of patience.

Hours later, I found the issue.

I fixed it.

And honestly, something changed in me after that night.

At first, I didn't think much of it. I just went back to doing what I normally did — writing backend code during the day, and quietly fixing infrastructure problems whenever they showed up. Around the same time, I started writing shell and Python scripts to automate deployments. Mostly because I got tired of repeating the same manual release steps over and over again. Nobody asked me to do it. It just felt wrong that we were wasting so much time doing things manually.

Then one day, my manager pulled me aside and said: “Why don't you do this full time?”

I remember being confused. I genuinely didn't know what he meant. That's when he introduced me to the idea that this was an actual career path — that there were people whose entire job was solving infrastructure problems, automating systems, and making deployments reliable.

That conversation changed everything for me.

Because the truth was, I didn't just enjoy that environment — I felt at home in it. The black terminal screen never felt limiting to me. It felt like where I belonged.

From there, things started moving fast.

Once “DevOps” became my actual job title, I went all in. I spent my days building Jenkins pipelines, writing Ansible playbooks, setting up Fastlane for mobile deployments — basically trying to automate every painful, repetitive thing teams hated doing.

My idea of success became simple: release day should feel boring.

And slowly, it did. We automated away nearly 70% of the manual release process, and deployments stopped being these stressful all-hands-on-deck events. The team could finally focus on building instead of firefighting.

Around that time, I also felt this pull to experience life outside India. I'd spent my whole life in one country, and I wanted to see more of the world. The UAE felt like the right first step. Familiar enough to not feel overwhelming, but different enough to completely change my perspective.

What surprised me most wasn't the move itself — it was the scale of the systems waiting for me there.

We were dealing with a logging platform ingesting around 600 GB of data every single day. A cloud gaming platform handling 450,000 requests per minute across 14 live games. Video delivery systems spread across AWS and Alicloud regions.

Nobody hands you documentation for problems like that.

You learn by sitting with the system long enough to understand how everything connects. At that scale, infrastructure problems stop being “just infrastructure.” They become systems problems. You can't fix one piece unless you understand the entire flow around it.

Those two years in Abu Dhabi completely changed the way I think about engineering.

Then came Berlin.

That move was personal and professional at the same time — the two were impossible to separate. I wanted the experience of living in Europe, but I was also excited by the kind of engineering challenges waiting there.

I worked on bringing structure to systems that had become messy over time — deployment pipelines, access management, observability. We reduced deployment effort by around 40%, which made life easier for everyone touching production.

Then things got even more interesting. I helped lead a migration from a monolith to microservices using cdk8s, writing Kubernetes manifests in TypeScript. In a strange way, it felt like my backend developer days coming full circle — except now I understood not just the application code, but everything happening underneath it too.

And then came the current chapter.

I wasn't hired to maintain existing infrastructure. I was hired to design validator infrastructure for Cosmos chains from the ground up. That meant building Helm chart libraries, implementing Horcrux threshold signing for validator security, and creating zero-downtime upgrade processes for live validators handling real blockchain transactions.

The stakes felt different here.

In most infrastructure jobs, a mistake might mean downtime or a failed deployment. Here, a mistake could mean missed blocks, slashed validators, and actual financial consequences on a public network.

It's probably the most precise engineering work I've done in my career.

And definitely the most fascinating.