DevOps Engineer focused on cloud infrastructure, Kubernetes, Infrastructure as Code, CI/CD, observability and automation, with a growing focus on cloud architecture, security, platform engineering and scalable production systems.
Cloud
Platform
Observability
Hey, I'm Prakash — a DevOps Engineer who genuinely enjoys taking messy, fragile infrastructure and turning it into something automated, observable and reliable.
Day to day, I'm usually somewhere in AWS, Azure, Kubernetes, Terraform, CI/CD or an observability stack. I like the kind of problem where a deployment that used to take an afternoon of manual steps turns into a five-minute pipeline run.
Right now I'm pushing further into cloud architecture, platform engineering, security and advanced Kubernetes — the stuff that turns "it works" into "it works reliably, at scale, without anyone getting paged at 2am."
A timeline of infrastructure roles, each expandable for the details.
Amazon Web Services
Opus
Anthropic
Not a logo wall — an honest read on where each platform sits: shipped in production, or actively being built up.
Cluster
Node A
Node B
Modules, not one-off scripts — environments separated deliberately, state managed remotely.
Reusable, versioned modules shared across projects — no copy-pasted resource blocks per environment.
Clean separation between staging and production, each with its own state and variables.
Remote state management so infrastructure changes stay safe under concurrent, collaborative work.
Validation and plan review built into the deployment pipeline, not run ad hoc from a laptop.
From commit to running in a cluster, with a monitoring loop back to the deploy that caused it.
Security treated as layers around the application, not a single gate before deploy.
A highly-available reference architecture. Hover any layer to trace the request path — or inject a failure below and watch it adapt.
Pick a scenario to see how I'd reason through it — requirements, architecture, scaling, security and the trade-offs that come with each.
Requirements
Scaling Strategy
Horizontal pod autoscaling on request latency and queue depth, fronted by a CDN to absorb read-heavy bursts.
Security
WAF at the edge, least-privilege IAM between services, network policies isolating the data tier.
Observability
Request-level tracing through the load balancer and services, with SLOs on p99 latency.
Cost Considerations
Cache hit ratio directly controls database load and compute spend — the first lever to pull.
Trade-offs
Caching improves latency but introduces staleness that has to be bounded deliberately.
Automation isn't the code — it's what the code lets me stop doing by hand. Select a system to see what runs underneath.
Sweeps EC2/ALB target health across a region and flags anything unhealthy.
Scans workloads across every namespace and flags anything stuck outside a healthy state.
Pulls daily cost-by-service from AWS Cost Explorer for trend tracking.
Treating infrastructure as a product for developers to self-serve — not a ticket queue.
If it isn't observable, it isn't operable — metrics, logs and traces feeding one place.
Cost as a first-class signal — visible, attributable and actionable across every provider.
Azure API Management as a centralized security and governance layer for Azure, on-premises and AWS-hosted applications.
Security & governance gateway
Hover for full policy list
Azure
Self-Hosted
Legacy App
On-Prem
Self-Hosted
Legacy App
AWS
Self-Hosted
Legacy App
Client → Legacy Application
Engineering Emphasis
A selection of highlighted projects from ongoing infrastructure work — not an exhaustive list.
Infrastructure as Code — AWS
Production and staging infrastructure provisioned through Terraform with environment separation and automated GitHub Actions deployments.
↓ Architecture
Centralized Multi-Cloud Cost Visibility
Centralized multi-cloud cost visibility and optimization platform.
↓ Architecture
Metrics, Logs & Traces Pipeline
Unified observability stack surfacing metrics, logs and traces from Kubernetes workloads into a single Grafana pane.
↓ Architecture
Legacy to Cloud-Native API Platform
Migration of legacy and on-prem APIs into a managed, automated Azure platform.
↓ Architecture
APIs Migrated
Years in Infrastructure / DevOps
Cloud Platforms
Production Environments
Infrastructure as Code
Growth in progress, not expertise already claimed.
Thoughts, experiences and observations from the cloud and DevOps community.
The real activity — repos, commits, everything — lives on GitHub @prakashgithub-un.
The areas I'm deliberately building depth in right now, edited here as they progress.