Hi, I’m Prazwal.
I make deploys
boring.
I build and operate the systems that keep software running at Gurzu — and I like to understand them all the way down to the kernel.
Good infrastructure is the kind nobody has to think about.
I keep the path from a developer’s commit to production short, boring, and reliable. Boring is the point — a deploy should be the least eventful part of anyone’s day.
I started in carrier networks, watching BGP routes at Worldlink and commissioning access networks at Huawei, so I still trace problems from the routing layer up. I’d rather understand a system all the way down than trust it as a black box.
The road from commit to production.
Zero-downtime deployments
Designed and shipped a health-checked rolling release process across our Kubernetes and CI/CD pipeline. No visible blip, no maintenance window — now the default way we ship.
Cloud for enterprise clients
Scalable infrastructure on GCP, AWS and Linode for clients with very different availability and compliance needs — shaped around each product, not one house pattern.
CI/CD pipelines
Built the team’s first pipelines with dev and ops, and keep them running across client environments — so a release is a pipeline run, not an event.
Monitoring & incidents
Services that report their own health instead of waiting for someone to notice. I lead root-cause response, then automate whatever keeps coming up.
Three things I optimise for.
Every pipeline, dashboard and runbook I touch gets measured against these.
Commit to prod, no meeting.
Pipelines trusted enough to run themselves. If shipping needs someone to babysit it, that’s a bug in the pipeline.
Notice it before the customer does.
Alerts that point at a cause, not a symptom. An alert nobody acts on is worse than no alert.
Delete the runbook.
Done twice, it gets written down. Done three times, it gets automated — and the doc gets deleted.
From the wire up to the pipeline.
Five years across telecom networks and cloud infrastructure — which is why I debug from the packet up.
Cloud infrastructure, CI/CD pipelines and production monitoring for enterprise clients on GCP, AWS and Linode.
Ran access network infrastructure for an internet-scale telecom customer, and installed and commissioned equipment in the field.
Watched MPLS, DWDM and BGP across NPIX links around the clock, flagging anomalies before they became outages.
What I reach for.
Grouped the way I actually use them — from bare Linux boxes to managed Kubernetes.
Notes from the layer below.
Things I only understood after reading the source instead of the docs — practical, with real commands.
What I’m up to.
- Running a zero-downtime release process for production at Gurzu.
- Studying for the CKA (Certified Kubernetes Administrator).
- Studying for AWS Certified Solutions Architect — Associate.
- Reading kernel networking code, one rabbit hole at a time.
Let’s make your releases boring too.
Hiring, scoping a piece of infrastructure work, or just want to trade notes on how things actually work underneath? I read every message.
Prefer plain email? bprazwal678@gmail.com