Rajan Verma · Mumbai
DevOps / CI platform engineer.
I’m a DevOps / CI platform engineer in Mumbai. Five years (Aug
2021–present) at TCS on the Ericsson account. I administer Jenkins
and Gerrit — admin panels, agents, access control, build servers,
and functional IDs — for a CI platform used by 300+ developers.
Day to day that means shared libraries, quality gates, and release
readiness: whether a change is allowed to land. I own platform
health in that sense (CI Guardian). I also own end-to-end CI for
the GTE module (15+ jobs), and I open-sourced Jenkins-XaaS — YAML
in, Jenkinsfile out, pipeline onboarding from hours to seconds.
Experience
DevOps / CI Platform Engineer
Tata Consultancy Services — client: Ericsson. Mumbai. Aug 2021 –
present.
-
Authored and maintained Jenkins Shared Libraries used as the
foundational CI framework across product and integration pipelines.
-
Built and open-sourced Jenkins-XaaS — YAML configs become
Jenkinsfiles, jobs deploy via the Jenkins REST API; pipeline
onboarding from hours to seconds.
-
Designed and maintained declarative Jenkins pipelines across 3 CI
repos, 10 product repos, and about 41 jobs (CI, product, and
utility) — build stability, quality gates, and release readiness.
-
Jenkins and Gerrit administrator — independently built and operates
the environment. Day-to-day: admin panels (plugins, agents, access
control), Linux build-server administration, and functional
(service) IDs used by CI jobs; recognised by the Ericsson CI
Architect.
-
Own CI platform health for 300+ developers — build stability,
quality gates, pipeline health, and release readiness (CI Guardian)
across a 65-member CI team.
-
Engineered parallel static analysis pipelines, cutting runtime from
300+ hrs to under 75 hrs/month — saving 200+ engineer-hours
monthly.
-
Built a self-healing Jenkins workflow to auto block/unblock Gerrit
submissions, automating a 12-person manual merge gate.
-
Automated CI monitoring across 12 Gerrit repos, reducing manual
monitoring from 60 mins to 5 mins per cycle (91% efficiency gain).
-
Python scripts for CI capacity planning and inline static-analysis
warning injection into Gerrit, reducing code-review cycles for
170+ developers.
-
GTE module CI owner — end-to-end CI for the GTE module, 15+ jobs;
worked with GTE module owners whose average engineering experience
exceeds 28 years, including contributors to the Erlang programming
language.
-
Flow Development Team — hiring (interviews), training, sprint
planning, and code reviews. Activity on the team, not a lead title
on paper.
-
Functional-ID lock monitoring Jenkins job — detect lock timestamps
and root causes, so the same CI disruption does not repeat.
-
Jenkins job tracker — query and report jobs triggered in a time
window for audit trails, incident analysis, and build history.
-
Primary SPOC for platform upgrade and migration — Gerrit migration
and Jenkins version upgrades end-to-end, including verification and
stakeholder communications (advance notice, window, completion) to
300+ users.
-
Migrated build servers from RHEL 6 to RHEL 7/8 — preliminary
testing and package validation with module teams; production
cutover within a 3-hour window.
-
Upgraded Gerrit from v2.14 to v3.6 — data integrity, performance,
and access governance.
-
Automated FOSS security scanning with BlackDuck and enforced
Nexus/Artifactory artifact retention across the CI ecosystem.
Open source
Jenkins-XaaS
Self-service CI/CD: YAML in, Jenkinsfile out, jobs deployed via the
Jenkins REST API. Pipeline onboarding from hours to seconds. Shared
library plus a CLI. GitHub Actions for that project’s job-deployment
orchestration. Public repo, Jul 2024–ongoing.
Skills
Git, Jenkins (pipelines, shared libraries, admin), Gerrit (admin),
Maven, Groovy, GitHub Actions. Docker. Linux (RHEL 6/7/8, CentOS).
Nexus, Artifactory, BlackDuck. Jira, Confluence, Eiffel.
Learning, not certifications held: AWS (prep for DevOps Professional).
Terraform basics. Kubernetes = basic concepts.
Education
Bachelor of Engineering — Information Technology. Shree L.R. Tiwari
College of Engineering. CGPA 8.03 / 10. 2017–2021.
AWS Certified DevOps Engineer – Professional — in progress (expected
Q4 2026). Not held.
Personal projects
On-prem inference, retrieval, and serving CI
Three public, labeled demos: an on-prem generate API, a private
retrieval API over a public-domain corpus, and a status page with
live health probes plus a written idle policy. How I practice
AI-infra skills. Not client work. Not a product.
Stack that actually runs: FastAPI; CPU Ollama (qwen2.5:1.5b);
Qdrant; Caddy reverse proxy and Let’s Encrypt HTTPS; Bearer auth.
No GPU bill. Not Ericsson CI. Each push to a private GitHub repo
runs GitHub Actions: CPU Ollama on ubuntu-latest, then health, a
rejected token, and a real generate and retrieve. That job does not
deploy this host — I copy a green commit here myself.