BIOS Version 2.4.1 · AMD Ryzen 3 2200G · 16384MB DDR4
─────────────────────────────────────────────────────────
[ kernel ] Loading Linux 6.8.0-generic... OK
[ init ] Starting systemd... OK
[ net ] Bringing up eth0 (192.168.0.112)... OK
[ docker ] Starting container runtime... OK
[ k3s ] Starting Kubernetes (chaos-creation)... OK
[ traefik ] Ingress controller ready... OK
[ grafana ] Metrics stack online... OK
[ vault ] Secrets engine unsealed... OK
[ wazuh ] SIEM agent active... OK
[ ansible ] Checking playbook integrity... OK
─────────────────────────────────────────────────────────
All systems operational. Uptime: 0d 0h 0m 1s
 
sai@chaos-creation:~$ ./portfolio.sh
$whoami
Sai_Kumar_Agiri
$cat /etc/role
devops engineer · release operations · infra
$uptime
5+ years keeping production alive

Sai.

I operate the infrastructure nobody sees — until it breaks.
5,000+ Linux servers. 9 data centers. Zero unplanned downtime.

CEH v10 CKA (in-progress) Pune, India

I'm a DevOps and Release Operations Engineer with 5+ years of experience automating large-scale Linux infrastructure across hybrid On-Prem and AWS environments. Currently at PubMatic, where I run CI/CD pipelines, Kubernetes deployments, and multi-DC release workflows at scale.

My work lives in the parts of the stack most people only notice when they catch fire — config management, container orchestration, deployment pipelines, incident response, security hardening. The invisible scaffolding that holds everything else up.

I start every system from first principles: no assumptions, no inherited mess. I'm also a Certified Ethical Hacker (CEH v10), which means I think about how things break as much as how they run.

When I'm not on-call, I run a fully automated homelab called chaos-creation — a bare-metal K3s cluster I use to learn, break things safely, and document what I find. You can read the full build walkthrough on Medium.

Cloud & IaC

AWS EC2 EKS VPC ALB IAM S3 Route 53 Terraform

CI/CD & Automation

Jenkins GitHub Actions GitLab CI Ansible SonarQube

Containers & Orchestration

Docker Kubernetes K3s EKS Helm Traefik

Observability

Prometheus Grafana Alertmanager ELK Stack Nagios

OS & Scripting

Linux (Ubuntu) CentOS RHEL Bash Python YAML

Networking

NGINX F5 Load Balancer VPN Firewall Rules VPC/Subnetting

Security & Compliance

HashiCorp Vault Wazuh SIEM Nessus OWASP CIS Hardening

Collab & Tooling

Git GitHub GitLab JIRA Confluence
chaos-creation homelab · personal infra
↗ github   ↗ read writeup

Fully automated homelab built on a single bare-metal AMD Ryzen 3 machine running Ubuntu 24.04. The entire stack — from OS prep to running services — is provisioned with Ansible playbooks, no manual steps. Runs a K3s cluster with Traefik ingress, a Grafana + Prometheus monitoring stack, Jellyfin media server, Uptime Kuma, and a Homepage dashboard. If the machine crashes, 9 playbook runs restore everything from scratch.

K3s Ansible Docker Prometheus Grafana Traefik Helm Jellyfin Ubuntu 24.04 Bare Metal
homelab · security layer in-progress · hardening

Next phase of chaos-creation focused on locking it down: Tailscale for zero-trust remote access, cert-manager + Let's Encrypt for HTTPS across all services, UFW firewall rules, Ansible Vault for encrypting secrets at rest, and Fail2ban for brute-force protection. Because a homelab without security is just a homelab waiting to be compromised.

Tailscale cert-manager Ansible Vault UFW Fail2ban Let's Encrypt
darkroomdev.in this site · hand-coded
↗ github

Personal site built from scratch — no frameworks, no CMS, no templates. Static HTML and CSS deployed via GitHub Pages with a custom domain. The way a sysadmin builds a website: minimal dependencies, fast, fully owned.

HTML CSS GitHub Pages Namecheap

Certifications

  • Certified Ethical Hacker — CEH v10  (ECC5420618973)
  • Certified Kubernetes Administrator — CKA  in-progress

DevSecOps & Tooling

  • Integrated HashiCorp Vault for dynamic secret injection across 100+ CI/CD pipeline jobs — eliminating hardcoded credentials across all environments
  • Embedded SonarQube SAST + Wazuh SIEM checks directly into Jenkins release pipelines for continuous security scanning at every deployment stage
  • OS hardening and baseline configuration across a 5,000+ node Linux fleet via Ansible, enforcing consistent CIS-compliant security posture across data centers
  • Vulnerability assessment using Nessus and OWASP methodology across infrastructure and application layers
  • Hands-on experience with ITSM incident workflows and ITIL processes for security-related incidents and P1/P2 response

Homelab Security (in-progress)

  • Tailscale zero-trust VPN for secure remote access without exposing ports
  • Ansible Vault for secrets management and encrypted credential storage
  • UFW firewall + Fail2ban for perimeter hardening and brute-force protection
  • TLS everywhere via cert-manager and Let's Encrypt

Interesting infra problem? Want to talk systems, security, or homelab builds? Reach out.