Skip to content
Hire me

Open to work now

Experience

One company, four years and seven months, and every layer of it.

I joined to write backend services and left running the infrastructure for the entire portfolio. At a services company that progression is not a promotion path, it is what happens when you are the person who says yes to the thing nobody owns.

Download CVAvailability and rates

Senior DevOps & Full-Stack Engineer

Feb 2022 – Aug 2026

TST Technology, Ahmedabad, India

A services company, which is the detail that explains the rest of this page. There was no platform team to hand things to and no single product to specialise in. Every client platform and every company product ran through the same infrastructure, and for most of those four years I was the person it ran through.

  1. 2022

    start

    Joined as a backend developer

    Node.js, Express, schema design

    Client platforms, writing services and designing schemas. The infrastructure was somebody else's problem and I wanted it to be mine.

    • Node.js
    • Express
    • MySQL
    • PostgreSQL
    • REST

  2. 2023

    year 2

    Took over the infrastructure

    AWS, Docker, CI/CD, first on-call

    Built the first pipelines, containerised the services, and took the pager. The first outage that was mine to fix happened this year, and it changed how I think about alerts.

    • AWS EC2
    • Docker
    • Jenkins
    • Nginx
    • PM2
    • CloudWatch

  3. 2024

    year 3

    Kubernetes and real scale

    EKS, ECS, Terraform, ArgoCD

    Designed and ran the EKS environment behind a platform serving millions of requests a day. Split another product into ECS services so one hot workload could not take the rest down. Everything moved into Terraform.

    • Kubernetes (EKS)
    • AWS ECS
    • Terraform
    • ArgoCD
    • Helm
    • Redis
    • Elasticsearch

  4. 2025

    year 4

    Release engineering across three stores

    iOS, Android, macOS

    Set up and ran the store accounts from first enrolment: team roles, app records, bundle IDs, signing certificates, API keys for automated upload, and the review submissions after. Tag a commit, get a build in front of testers.

    • Xcode Cloud
    • App Store Connect
    • Google Play Console
    • notarytool
    • Fastlane-style CI

  5. 2026

    now

    Senior DevOps and Full-Stack Engineer

    Sole or lead infra engineer across the portfolio

    Owning infrastructure across every client platform and company product at once, and lead backend on most of them. Alongside it, building and running Anonymous India, which is entirely mine.

    • GCP
    • Socket.io
    • NestJS
    • Next.js
    • React
    • Flutter
    • Prisma

What I actually owned

Infrastructure and platform

  • Architected and ran EKS clusters for high traffic consumer platforms, including cluster design, node groups, autoscaling and rollout strategy.
  • Ran ECS microservices for a platform with 100,000+ users, split into separate task definitions so one service scaling could not exhaust shared capacity.
  • Terraform for everything, including the things we swore we would only do once.
  • Query latency and sustained load across PostgreSQL, MySQL, Redis and Elasticsearch. Caching strategy, replication and query tuning.
  • Linux estate: Nginx reverse proxy, SSL termination, load balancing, log rotation, hardening and remote execution.

CI/CD and automation

  • Delivery pipelines across Jenkins, ArgoCD, CircleCI and AWS CodeBuild or CodePipeline, matched to how each project is hosted rather than forced into one standard.
  • Cut rollback from a manual scramble to a single reverted commit by moving Kubernetes delivery to GitOps with ArgoCD.
  • Declarative, version controlled, auditable. A bad release stops being an event.

Mobile and realtime

  • Flutter across a large portfolio of apps, on a consistent toolchain: GetX for state, Dio for the API layer, and Firebase for auth, push and crash reporting.
  • Socket.io on both ends. Server side for live prices, leaderboards and notifications, and socket_io_client in the apps consuming them.
  • Push delivery end to end: Firebase Cloud Messaging, local notifications, permission handling and the token lifecycle that quietly breaks when nobody owns it.
  • Media pipelines behind the apps: Multer uploads, Sharp resizing, S3 storage and CloudFront delivery, so a content-heavy screen still loads on a bad connection.

Release engineering

  • Android release cycles end to end through Google Play Console: signing, versioning, staged rollout and store compliance.
  • iOS releases through App Store Connect, with Xcode Cloud handling automated builds and TestFlight distribution.
  • macOS release builds: universal binaries for both Apple Silicon and Intel, Developer ID signing, notarisation and stapling.

Backend and full-stack delivery

  • Backend development across the portfolio: Node.js, Express and NestJS services, REST APIs, WebSocket and Socket.io services, and schema design across PostgreSQL, MySQL and MongoDB.
  • Architecture across all three shapes: monolithic where it fit, microservices where a workload needed to scale alone, and event-driven with SQS, Lambda and Redis pub/sub for anything that did not belong on the request path.
  • AI features in shipped products, not demos: prompt-based search over a catalogue via the OpenAI API, and card recognition that turns a photographed business card into a contact.
  • Next.js and React front ends, so technical scope of a request runs from requirements to rendering.
  • Security as part of the build rather than a later pass: JWT auth, bcrypt, Helmet, per-route rate limiting and role based access, applied consistently across the portfolio rather than per project.
  • Client delivery directly, not through a layer. Urgent releases and production fixes outside business hours, often in languages and codebases I did not write.

The part that is not a tool

The half of this job that has no logo.

I have led delivery, not just my own tickets.

Owned client relationships directly and run the team through to launch. At a services company the person who can talk to the client and fix the pipeline is the same person, and that is usually me.

Breaking something is not a meeting.

I have sat with junior engineers after an incident instead of escalating it. People who are afraid of the deploy button ship worse things, more slowly, and hide it for longer.

You cannot automate a broken process.

You just get broken faster, with better logging. This took me the longest to learn. Now I sit with the process before I automate it, which is the boring half and where the real wins are.

I land the release when it has to go out.

Outside hours, in whatever language the problem happens to be written in, often in a codebase I did not write. That is not a boast, it is just what the job has been.

Nobody notices infrastructure when it works. I made peace with that a long time ago.

The stack, grouped honestly

Everything here has been in production on something I was responsible for. Nothing on this list is from a tutorial.

Cloud
  • AWS
  • Azure
  • GCP
  • EC2
  • EKS
  • ECS
  • ECR
  • RDS
  • S3
  • CloudFront
  • Lambda
  • SNS
  • SES
  • SQS
  • IAM
  • CloudWatch
Containers and orchestration
  • Docker
  • Docker Compose
  • Kubernetes
  • Helm
  • ArgoCD
  • HPA
  • Cluster Autoscaler
Infrastructure as code
  • Terraform
  • Nginx
  • PM2
  • systemd
  • Bash
CI/CD
  • Jenkins
  • ArgoCD
  • CircleCI
  • AWS CodeBuild
  • CodePipeline
  • Azure Pipelines
  • GitHub Actions
  • Xcode Cloud
Backend
  • Node.js
  • Express
  • NestJS
  • Fastify
  • TypeScript
  • REST
  • WebSockets
  • Socket.io
Data and ORMs
  • MySQL
  • PostgreSQL
  • MongoDB
  • Redis
  • Elasticsearch
  • Sequelize
  • TypeORM
  • Prisma
  • Mongoose
Queues and async
  • Amazon SQS
  • AWS Lambda
  • Redis pub/sub
  • node-cron
  • Scheduled workers
Media pipeline
  • Multer
  • Sharp
  • S3
  • CloudFront
  • Image resizing
Front end
  • Next.js
  • React
  • TypeScript
  • Tailwind
Mobile
  • Flutter
  • Dart
  • GetX
  • Dio
  • Firebase Messaging
  • Local notifications
  • socket_io_client
AI in production
  • OpenAI API
  • Prompt-based search
  • Image and card recognition
Security
  • JWT
  • bcrypt
  • Helmet
  • Rate limiting
  • RBAC
  • IAM policy
  • Secrets management
Observability
  • Prometheus
  • Grafana
  • CloudWatch
  • Alert tuning
  • Runbooks
Firebase
  • Auth
  • Cloud Messaging
  • Storage
  • Crashlytics
  • Analytics
  • App Check
Integrations
  • Stripe
  • Razorpay
  • PayPal
  • Msg91
  • Twilio
  • Nodemailer
  • RevenueCat
Release
  • App Store Connect
  • Google Play Console
  • TestFlight
  • notarytool
  • Code signing

Before that

Certification

Professional Web Development

Shree Academy, Rajkot

Six month programme, completed December 2021, graded A. This is where the foundation actually got built.

Education

Instrumentation & Control Engineering

Vishwakarma Government Engineering College, Chandkheda

2018 to 2022, not completed. I left to take the development work full time, and I would make the same call again.

Want the parts that do not fit on a page?

Happy to walk through any of it properly, including the decisions I would make differently now.

deepboda18@gmail.comSee the work