Automate the toil
Replace manual runbooks with GitOps, pipelines, and IaC that scale with the team.
DevOps Engineer · ap-south-1 /
Bengaluru
NITESH KUMAR
I own the full delivery lifecycle for 52+ services across 5 Kubernetes clusters — CI/CD, GitOps, observability, security, and cost, all built to hold up at 3 a.m.
status
open to opportunities
region
ap-south-1 · Bengaluru
since
2024 · 2y+ uptime
local
00:00:00
DevOps Engineering
AWS
Kubernetes
Amazon EKS
Docker
Helm
ArgoCD
GitOps
Terraform
Ansible
Infrastructure as Code
Jenkins
GitHub Actions
CI/CD Pipelines
Prometheus
Grafana
Loki
Alertmanager
OpenTelemetry
Coralogix
Observability
HashiCorp Vault
Trivy
Veracode
AWS WAF
DevSecOps
SOC 2
Apache Kafka
Amazon MSK
AWS EC2
AWS RDS
AWS ECR
AWS Athena
VPC
IAM
Cost Optimization
Incident Response
Python
Bash
01 — whoami
Automation isn't a task.
Automation isn't a task.
It's the job.
Two years making 3 a.m. pages rare: GitOps that self-heals, deploys nobody has to babysit.
I build and run cloud-native delivery platforms on AWS and Kubernetes — owning Dockerization, Helm, CI/CD, GitOps, secrets, observability, security & compliance, cost, and incident response for 52+ production services across Python, Java, and Node.js, with a backend foundation in Java and Spring Boot.
The throughline: automate the toil, secure by default, and own it end to end in production.
Secure by default
Shift-left scanning, Vault-managed secrets, and compliance baked into the delivery path.
Own it in prod
From root-causing OOM crashes to leading incident response and cutting MTTR.
02 — git log --career
Deployment history.
release/2025.03 — HEAD
Lumber
Mar 2025 → present
DevOps Engineer now
Solely own the full delivery lifecycle for 52+ production services across 5 Kubernetes clusters (Python, Java, Node.js) — Dockerization, Helm, CI/CD, and secrets. Drove org-wide GitOps with ArgoCD, migrated secrets to a HashiCorp Vault cluster, architected observability (Prometheus, Grafana, Loki, OpenTelemetry, Coralogix), hardened security & compliance (Trivy, Veracode, AWS WAF, SOC 2), cut ~$4.8K/mo in AWS spend, and led incident response to keep MTTR low.
Kubernetes
ArgoCD
AWS
Terraform
Vault
Prometheus
OpenTelemetry
Jenkins
Trivy
AWS WAF
release/2024.08
Lumber
Aug 2024 → Feb 2025 · 7m
Software Development Engineer — Intern
Optimized API performance by profiling and eliminating bottlenecks with YourKit, cut data storage footprint 13% by removing duplicate Postgres audit entries, and drove unit-test coverage past 93% across a multi-module Spring Boot application.
Java
Spring Boot
PostgreSQL
JMeter
YourKit
02a — git diff --stat
52+
Production services owned
27%
Faster CI builds
$4.8K
Monthly AWS spend cut
44K+
Malicious requests blocked / mo
Two years, diffed.
Production ownership, squashed into numbers.
03 — cat stack.yaml
The toolchain, declared.
production toolkit
Built over two years solving delivery, observability, and security challenges across 52+ services and 5 Kubernetes clusters.
1stack:2 cloud: [aws]3 containers: [docker, kubernetes, eks, helm]4 delivery: [argocd, jenkins, github-actions]5 iac: [terraform, ansible]6 observability: [prometheus, grafana, loki, alertmanager, opentelemetry, coralogix]7 data: [kafka, msk, postgres, redis]8 security: [vault, trivy, veracode, aws-waf, soc2]9 scripting: [python, bash]1011mode: automate_the_repetitive # gitops • secure by default
ArgoCD, Jenkins, GitHub Actions, Terraform, Ansible
AWS, Kubernetes, Amazon EKS, Docker, Helm
Prometheus, Grafana, Loki, OpenTelemetry, Vault, Trivy, AWS WAF
04 — ls -la credentials/
Proof, not promises.
certifications
CNCF
Certified Kubernetes Administrator (CKA)
The Linux Foundation
impact across the stack
Solely own the full delivery lifecycle for 52+ production services across 5 Kubernetes clusters (Python, Java, Node.js) — Dockerization, Helm charts, manifests, CI/CD wiring, secrets, and observability, end to end.
Drove org-wide GitOps by migrating every service to ArgoCD — automated sync, self-heal, drift detection, and one-click rollbacks with full deployment audit history.
Migrated secrets from Ansible playbooks to a HashiCorp Vault cluster using AppRole auth in a hub-and-spoke model — centralizing storage, access control, and rotation across every environment.
Architected end-to-end monitoring with the kube-prometheus-stack (Prometheus, Grafana, Loki, Alertmanager) and OpenTelemetry tracing; unified logs, metrics & traces in Coralogix, whose CSPM module surfaced 60+ security findings.
Engineered CI/CD for Python, Java & Node.js running unit, integration, and Playwright tests at PR time; parallelized Jenkins stages to accelerate builds 27%, and killed Docker Hub 429s with an ECR pull-through cache.
Embedded Trivy image scanning and Veracode SAST/DAST across every pipeline (shift-left DevSecOps), remediating 25+ CVEs to harden supply-chain security.
Solely drove infrastructure SOC 2 Type 1 & 2 — remediating weak TLS/CBC ciphers flagged by testssl.sh — and deployed AWS WAF at the edge, blocking 44K+ malicious requests/month.
Cut ~$4,800/month in AWS spend — Athena partitioning (99% cut), single-AZ non-prod RDS/EKS, and automated off-hours shutdowns — with org-wide tagging and billing alerts.
Migrated all 52+ services from AMD64 to ARM64 (AWS Graviton) with multi-platform Docker builds, and architected a fully isolated AWS environment (dedicated VPC, EKS, subnets, IAM, RBAC) for a client's network-segregation requirements.
Killed recurring production OOM crashes by provisioning PVCs to capture JVM heap dumps on crash, and led incident response — storage shortages, bastion crashes, zero-downtime node-group migrations, subnet/IP exhaustion — cutting MTTR.
education
Bachelor of Technology
Computer Science & Engineering · CGPA 8.1
HKBK College of Engineering