open to devops roles ✦ thane, mh

hi there, I'm 🌿

PragatiSingh.

mainframe survivor ✦ cloud romanticist ✦ AWS in progress ☁
"She is the one who create herself."

6 years in IT. Started with COBOL and JCL at Cognizant & TCS, built real enterprise banking systems for some of the world's largest financial institutions. Then one night, discovered DevOps and everything changed. Now I'm building cloud infrastructure, containerizing things, and deploying dreams to production ☁

👤
your photo here ✿
replace with <img src="photo.jpg">
open
to
hire ✦
☁ AWS
🐳 Docker
⚙️ Terraform
☸ K8s
☁ deploying dreams to production 🐳 tiny containers, huge emotions ⚙ terraform: push & pray 🐧 surviving on grep & caffeine ☁ financially supporting AWS 🔁 ci/cd or cry/cd 📦 helm charts & heartache 🐛 features that became bugs ☁ deploying dreams to production 🐳 tiny containers, huge emotions ⚙ terraform: push & pray 🐧 surviving on grep & caffeine

📌 pinned note to self

Who is Pragati?

✦ ✦ ✦ ✦ ✦
The quietest engineers are often building the loudest systems.

I am a calm, grounded thinker who spent 6 years in enterprise IT — writing COBOL, building APIs, deploying mainframe systems for some of the biggest banks in the world. And while doing all of that, I was secretly becoming a DevOps engineer at night.

B.Tech in Electronics & Communication Engineering from Galgotia College of Engineering, Greater Noida. Graduate of 2019. Currently a Systems Engineer at TCS — working in big banking projects — and building my cloud skills every single day.

No bootcamp. No mentor. Just documentation, YouTube, broken pipelines, and a lot of "why is this still not working" at 1am. That is the whole story.

🌱 Deep learner — goes to the root before touching the code
🎯 Disciplined — mainframe background built precision habits
✦ Self-reliant — no shortcuts, just grit and documentation
☕ Calm under pressure — debugging feels meditative to me
🔍 Honest about what I know and what I am still learning
6
years in IT
6+
GitHub projects
💯
self-taught
☁ cloud & infrastructure
AWS (EC2, S3, VPC, Lambda)75%
Terraform65%
Docker80%
Kubernetes60%
🔁 devops tools
GitHub Actions / CI-CD72%
Linux & Bash Scripting78%
GIT & Version Control80%
🖥 mainframe (professional)
COBOL / JCL / VSAM / DB292%
CICS / API Building / Swagger85%

💼 career receipts

Places that paid me

✦ ✦ ✦ ✦ ✦
Dec 2022 — Present ✦ current
Systems Engineer
Tata Consultancy Services (TCS) ✦ Thane, Maharashtra
Working on enterprise banking systems for a leading global financial institution. Built COBOL programs for balance mismatch detection — fetching account balances via cursor, comparing three-way values, generating reports for business teams. Also worked on instant cashback feature for debit cards: batch processing daily transactions, calculating cashback points, applying minimum spend thresholds, and building APIs so customers can view cashback via the mobile app (SAR/AAR/Swagger files, tested via SoapUI).
COBOLJCL DB2API Building SwaggerSoapUI TCS
Nov 2019 — Dec 2022
Associate Developer
Cognizant Technology Solutions ✦ Chennai
Worked for a leading international financial services group. Part of the development, testing, and automation team. Built and maintained APIs and services, handled production deployments, maintenance tasks and enhancement projects. Used Z/OS Connect for mainframe API integration, ESB for messaging, JIRA for project tracking, and Postman for API testing.
COBOLJCL VSAMDB2 CICSZ/OS Connect PostmanESB Cognizant
2015 — 2019 ✦ 📚 education
B.Tech in Electronics & Communication Engineering
Galgotia College of Engineering & Technology ✦ Greater Noida, UP
Engineering graduate — built the foundation of system thinking, networking, and problem solving that now makes cloud concepts click faster. The discipline of engineering school + the precision of mainframe work = a DevOps engineer who actually understands why things work (and why they break).
B.Tech ECE Engineering Graduate Greater Noida

☁ a very honest story

From COBOL to Cloud

✦ ✦ ✦ ✦ ✦
I did not plan any of this. I came from mainframes — COBOL, JCL, batch jobs at 3am. One evening I watched a video about Docker containers and thought: wait, this is how modern systems should work. That one thought changed everything. Here is the real, unfiltered version of how I went from Z/OS to Kubernetes, one terminal session at a time.
01
2019–2022 ✦ the grind era
Mainframe life begins
Started at Cognizant, then TCS. Learned to think in systems — batch jobs, cursor logic, enterprise-grade precision. COBOL taught me that bugs in production are not jokes. This phase made me disciplined.
COBOLJCLDB2CICSAPI Building
02
2023 ✦ the spark
One Docker video at 11pm
Stumbled onto a Docker tutorial. One video became ten. Opened AWS free tier at midnight on a Wednesday. Created my first EC2 instance. It was just a tiny Linux server — but it felt like magic. I have not stopped since.
Docker basicsAWS free tierLinux terminal
03
2023–2024 ✦ the painful learning
Self-teaching everything from scratch
No bootcamp. No mentor. Just documentation and YouTube. Learned Linux deeply — file permissions, shell scripting, process management. Broke things. Fixed them. Broke them again. Kept going. "Permission denied" became almost a greeting.
LinuxBashAWS EC2/S3/VPCGit
04
2024–2025 ✦ building for real
Real projects. Real pain. Real growth.
Built a system performance monitoring tool. Set up CI/CD pipelines. Deployed web apps on EC2. Created Docker projects. All pushed to GitHub. Not just following tutorials — actually building. Every project taught me something tutorials could not.
DockerAWS EC2GitHub ActionsCI/CDShell Scripting
✦ meanwhile, in my DevOps life ✦
☁
CI/CD section
push to main and pray
🐳
Docker section
tiny containers, huge emotions
🐧
Linux section
surviving on grep & caffeine
💸
AWS section
financially supporting amazon web services
🐛
bugs section
features that accidentally became bugs
⚙
terraform section
currently containerizing my life decisions

🛠 things i built instead of resting

Real Projects

✦ ✦ ✦ ✦ ✦

all real. all on github. all built from scratch. some worked first try (lies). click any project to visit the repo ✦

★ main project · built end-to-end with DevOps
🌐
Docker GitHub Actions AWS EC2 Shell
My DevOps Portfolio — the site you're looking at right now ✿
This portfolio website is itself a DevOps project. Built the full deployment pipeline from scratch — containerized with Docker, version-controlled with Git, automated with GitHub Actions CI/CD, and deployed live on AWS EC2. The site you're reading right now was shipped using the exact tools I'm telling you I know. That's the whole point.
problem → solution → impact ↓
✦needed to prove DevOps skills, not just list them on a page
✦built and shipped the portfolio using the real DevOps toolchain end-to-end
✦fully automated pipeline · deployed on AWS · verifiable on GitHub
⟶ github repo 🌐 live site (you're here)
how it gets deployed ↓
git push
→
GitHub Actions
→
Docker Build
↓
AWS EC2
←
Docker Push
←
Registry
↓
🌐 live & running ✓
⟶ view on github HTML · CSS · Docker · GitHub Actions · AWS
☸ kubernetes gitops · two repos · one platform
☸
Kubernetes GitOps Docker Helm
Kubernetes GitOps Platform — config + platform, one complete system ✦
Two GitHub repos, one platform. The config repo holds all declarative cluster configs — Helm charts, Kubernetes manifests, environment-specific values. The platform repo handles the infrastructure layer — cluster provisioning, operator setup, tooling. Together they form a full GitOps workflow where Git is the single source of truth for everything running in the cluster.
problem → solution → impact ↓
✦manual kubectl deploys are error-prone, hard to audit, unscalable
✦GitOps model — all changes go through Git, auto-synced to cluster state
✦declarative · auditable · self-healing · zero manual kubectl in production
⟶ config repo ⟶ platform repo
gitops flow ↓
git commit
→
PR merged
→
GitOps sync
↓
k8s cluster
←
Helm apply
←
config repo
↓
☸ reconciled & live ✓
⟶ config repo ⟶ platform repo
Kubernetes · Helm · GitOps · YAML
📊
Shell Linux
System Performance Monitoring Tool
A bash-scripted tool that monitors CPU usage, RAM usage, and storage usage of a server in real-time. Includes backup-and-cleanup scripts. Real DevOps work — the kind that actually runs on prod servers.
how it works ↓
Server
→
Bash Script
→
Monitor CPU
RAM Storage
→
Alert / Report
→
Cleanup
⟶ view on github Shell 100%
🔁
CI/CD Multi-env
Multi-Environment CI/CD Pipeline
Implements a complete multi-environment CI/CD pipeline — dev, staging, and production environments with automated build, test, and deploy stages. Push to main and it just works (mostly).
pipeline flow ↓
Code Push
→
Build
→
Test
→
Dev → Staging
→
Production ✓
⟶ view on github CI/CD
🐳
Docker
Docker Projects Collection
A dedicated repository of Docker assignments and projects — containerizing applications, writing Dockerfiles, managing images and containers. Every project here came with a learning (and usually a headache).
docker lifecycle ↓
Dockerfile
→
docker build
→
Image ✓
→
docker run
→
Container 🐳
⟶ view on github Docker
🐧
Shell Linux
Linux Zero to DevOps
Collection of Linux and Bash scripting solutions with real-world DevOps use cases and interview-focused implementations. Built this as I learned — raw, honest, and incredibly useful.
learning path ↓
Zero Linux
→
File System
→
Bash Scripts
→
DevOps Ready
⟶ view on github Shell 100%
☁
AWS JavaScript
Web App Deployment on AWS EC2
Deployed a full web application on an AWS EC2 instance. Set up the instance, configured security groups, installed dependencies, and got the app running live on the cloud. My first real cloud deployment.
deployment flow ↓
AWS Console
→
Launch EC2
→
SSH In
→
Deploy App
→
Live ✓
⟶ view on github JavaScript
📦
Docker JavaScript
Docker Projects — Advanced Experiments
More Docker projects — multi-container setups, networking between containers, volume management, and more complex containerization experiments. Where the real Docker education happened.
multi-container setup ↓
App Container
⇄
Docker Network
⇄
DB Container
⟶ view on github JavaScript

🌱 the real curriculum

Oops & Lessons

✦ ✦ ✦ ✦ ✦

most portfolios show the wins. this one shows everything. because the mistakes are where the actual learning lives. these are real, raw, and mine. 🌿

🔥
devops fail ✦ aws
I accidentally made an S3 bucket public for 45 minutes
Was setting up a static site on AWS. Toggled the wrong setting. Realized 45 minutes later that some test files were publicly accessible. My heart has never beaten faster in my life.
💡 Lesson: Enable S3 Block Public Access at the account level FIRST. IAM policies and bucket policies are not the same thing. Always.
💥
devops fail ✦ funny & painful
I destroyed the wrong CI/CD environment
Was cleaning up a dev pipeline setup. Ran the wrong teardown command on what I thought was dev. It was not dev. Two weeks of config: gone. I sat very still for a long time after that.
💡 Lesson: Always print your current workspace/environment before any destructive command. Read the plan output three times. Then read it again.
🤐
career mistake ✦ personal growth
I waited almost a year to push my work to GitHub
For nearly a year I was building things but keeping everything local. I thought my work was not "good enough" yet. I was wrong. Perfectionism is just fear with better PR.
💡 Lesson: Done and visible beats perfect and hidden. Push to GitHub even when it's messy. You learn faster when others can see the work.
😔
personal growth ✦ mindset
I compared my journey to people with mentors and bootcamps
For a while I felt behind. Everyone seemed to know more, move faster, have more support. I was self-teaching from YouTube and documentation. It felt lonely and unfair some nights.
💡 Lesson: Self-taught is not slower — it is deeper. You understand the "why" because you had to find it yourself. That is your actual unfair advantage.
🐳
technical mistake ✦ docker
My Docker container kept restarting and I had no idea why
Spent an entire weekend debugging a container that kept crashing on startup. Checked logs a hundred times. The issue? One ENV variable was missing. One single missing variable. The logs never clearly said so.
💡 Lesson: Always check ENV variables and run docker logs with --follow first. Then check the Dockerfile. Then cry. Then check ENV variables again.
🌱
funny mistake ✦ linux
I ran rm -rf in the wrong directory
I was cleaning up some test files on a lab server. I was in the wrong directory. Gone. All of it. Just a clean, empty directory where my project used to be. I learned about backups that day.
💡 Lesson: Always pwd before rm. Always. Non-negotiable. Also: backup scripts exist for a reason (I now build them as projects).

✍ pages from my diary

Tiny Thoughts

✦ ✦ ✦ ✦ ✦

some things I'm okay sharing with the world. other things are just mine — written for myself first. if a blog is locked, you can ask me. I'll decide if I want to let you in. no hard feelings either way 🌿

☁ devops & learning
Why I chose DevOps when no one around me understood what it even was
When I told people I was learning cloud infrastructure, most nodded politely and changed the subject. Here is why I kept going anyway...
Apr 2025
💼 career
What mainframe actually taught me about DevOps (it's a superpower)
Everyone assumes mainframe is dead-end thinking. I think it gave me something most cloud engineers don't have — precision under pressure...
Mar 2025
🐳 technical
The honest beginner's guide to Docker — written by someone who broke everything first
No gatekeeping. Just real talk about what Docker is, why it hurts at first, and why it starts making beautiful sense...
Feb 2025
🔒 private blog
The night I cried because a deployment failed and I didn't know why
It was 2am. Everything had been working fine and then suddenly nothing was. I had no answers and I just...
🚫 sorry, you don't have access to read this.
this is a personal blog. ask Pragati for access — she'll decide if she wants to let you in.
🔒 private blog
What it actually feels like to be a woman self-teaching tech in India
Nobody talks about this part. The doubt that comes not just from yourself but from the people around you who mean well but...
🚫 sorry, you don't have access to read this.
this is a personal blog. ask Pragati for access — she'll decide if she wants to let you in.
🔒 private blog
The afternoon I almost gave up on switching to DevOps entirely
There was one specific afternoon where I had applied to so many roles and heard nothing back. I sat with my laptop closed...
🚫 sorry, you don't have access to read this.
this is a personal blog. ask Pragati for access — she'll decide if she wants to let you in.