DevOps engineer · Software engineer · Freelancer
Reliable delivery.
Calmer releases.
I’m Sourabh Dey. I build the pipelines, cloud infrastructure and software that let teams ship often and sleep well, from first deployment to production operations.
Services
Everything between a commit and a healthy production.
Eight focused service lines, delivered by one engineer who understands both the code and the platform it runs on.
01CI/CD pipelines
Automated build, test, scan and deploy workflows, so every release is repeatable and a bad change is stopped before it reaches users.
- Pipeline-as-code for new and existing apps
- GitLab CI/CD, Jenkins, GitHub Actions
- GitOps delivery with Argo CD and Flux
- Quality gates: tests, linting, image scans
- Environment promotion and approvals
- Rollback and release strategies
02Cloud infrastructure (AWS, Azure, GCP)
Secure, cost-aware cloud environments designed for the workload, from a first deployment to a migration off a legacy host.
- Architecture and landing-zone design
- Compute, networking, storage and IAM
- Migrations and re-platforming
- High availability and backups
- Security hardening and least privilege
- Cost review and right-sizing
03Docker and Kubernetes
Containerised applications and the platform they run on: clusters, ingress, storage, scaling and day-two operations.
- Dockerfiles and image build optimisation
- Kubernetes and OpenShift deployments
- Helm charts and manifests
- Ingress, load balancing, persistent storage
- Cluster upgrades and troubleshooting
- Workload security basics
04Infrastructure as code (Terraform, Ansible)
Environments described in version-controlled code, so they can be reviewed, rebuilt and kept consistent across teams.
- Reusable Terraform and OpenTofu modules
- Remote state, locking and plan review
- Ansible roles and configuration management
- Automated environment creation
- Drift detection practices
- Documentation your team can own
05Monitoring and logging
Visibility into what is actually happening: metrics, dashboards, central logs and alerts that tell you something useful.
- Prometheus and Grafana setups
- Central log search with Loki
- Dashboards for health and saturation
- Alert rules tied to user impact
- On-call routing with Alertmanager
- Review of noisy or missing alerts
06Site reliability
Practices that make incidents rarer, shorter and calmer: objectives, runbooks, capacity thinking and learning from failures.
- Service-level indicators and objectives
- Runbooks and incident checklists
- Backup and restore testing
- Capacity and failure-mode review
- Post-incident reviews
- Reliability roadmap for your team
07Automation
Removing repetitive operational work with scripts, tooling and workflows that are tested and documented.
- Python and Bash tooling
- Scheduled jobs and operational workflows
- Backup and housekeeping automation
- Provisioning and onboarding automation
- Internal CLIs and small services
- Integration with chat and ticketing
08Software development
Maintainable backend and web software: APIs, dashboards, internal tools and websites, built so they are easy to deploy and run.
- REST APIs and backend services
- Front-end and responsive websites
- Database design and integration
- Internal tools and dashboards
- Deployment-ready packaging
- Performance and SEO basics
Selected work
Reference builds that show how I work.
Open any card for the architecture, the approach and the technology behind it.
How I work
Good engineering is what you see. Care is what I bring.
- 01
Automation first
Anything done by hand more than twice becomes a script, a pipeline or a template that is tested and documented.
- 02
Security-conscious by default
Least privilege, secrets kept out of code and sensible hardening from the first commit, not as an afterthought.
- 03
Maintainable and documented
Readable code and short, honest documentation so you or any other engineer can pick it up later.
- 04
Clear scope and communication
Plain-language updates, no jargon-walls, and no guaranteed-uptime or guaranteed-savings promises I cannot stand behind.
Process
A better process, without the complexity.
Step 1: Discover
We talk through your goals, current setup, constraints and budget, and I ask the questions that surface the real problem.
Step 2: Design
I propose an approach, architecture and milestones in writing, so scope and cost are agreed before work starts.
Step 3: Build and deploy
I implement, automate and test, then ship in small steps with each change reviewed and reversible.
Step 4: Support and improve
Handover documentation, monitoring and optional ongoing support, then iteration as your needs change.
Technology stack
The tools I work with.
Grouped by purpose. Hover, focus or tap a tool to see what I use it for. No fake proficiency bars, just the stack.
Brand glyphs from the Simple Icons project (CC0). Where no glyph exists the tool is shown with a monogram. Trademarks belong to their owners.
How delivery works
From commit to production, interactive.
Select a stage to see what happens there, or run the simulated deployment to watch a change flow through the pipeline.
About
Engineering reliable systems. Building better software.
I’m a DevOps and software engineer working remotely from India. I help teams build, deploy and operate digital products: pipelines that remove manual release steps, infrastructure that can be rebuilt from code, and applications that are straightforward to run.
- Automation-first engineering
- Reliability before novelty
- Clean, maintainable implementations
- Security-conscious design
- Practical problem-solving
- Clear, direct communication
Contact
Have a project in mind? Let’s talk.
Tell me what you’re building, what needs improving, or where your infrastructure is getting in the way. I’ll reply with a practical next step.
FAQ
A little clarity. A lot of confidence.
What kind of work do you take on?
DevOps and cloud work (CI/CD, Kubernetes, Terraform, monitoring, reliability) and software development (APIs, internal tools, websites). Projects can be one-off, short engagements or ongoing support.
Do you work with existing infrastructure?
Yes. Most of the time the first step is understanding what is already running and improving it safely, rather than starting again.
How do we get started?
Send a short description through the form, WhatsApp or email. I reply to discuss goals, constraints and timeline, then propose scope and cost in writing before any work begins.
Which time zone do you work in?
India Standard Time (UTC+5:30), working remotely. I schedule calls to overlap with your working hours where practical.
Are the projects on this page client work?
No. They are labelled demo projects: illustrative reference builds that show how I approach each type of engagement. Real case studies will replace them as they can be shared.
Is the delivery pipeline demo real?
It is a browser simulation. It illustrates the stages of a typical delivery workflow and does not touch any real server, registry or cluster.