Sai Chand
Boxes moving along a small belt

Services

The work side

The notebook comes first. This page is for when you want a person to look at a cluster, a pipeline, or a cloud account and write down what is actually going on. No prices. If it fits, we start with a conversation.

01

Kubernetes setup, operations, and troubleshooting

A cluster the team can deploy to, and a way to find the layer that dropped the request.

The problem
The cluster is fine on a quiet Tuesday and a guessing game the day something times out. Or the team has outgrown one box and needs a platform they can operate without folklore.
What I do
I set up a cluster, or sit with the one you already have. Namespaces, workloads, ingress, DNS, storage, limits, and network policies. When a request dies, I trace it from the client back to the pod and name the layer that failed.
What you get
A cluster your team can ship to, notes from the failures we actually hit, and a short runbook for the paths that wake people up.

02

CI/CD and GitOps pipelines

A path from a merge to a running workload, with a record of what was built.

The problem
Releases depend on someone remembering the steps. Or the pipeline is green and the cluster is still running last week's image.
What I do
I build the path in Jenkins or GitHub Actions: build, test, scan, push. Argo CD (or the same idea in another GitOps controller) makes the cluster match git. Where they earn their keep, I add SonarQube, Trivy, and an SBOM stored with the artifact.
What you get
A merge that produces an image, a git revision the cluster is meant to match, and a way to roll it back without a treasure hunt.

03

Infrastructure as Code with Terraform

Infrastructure you can plan, review, and recreate.

The problem
The cloud account is a pile of click-ops. Or the Terraform you have is one state file nobody wants to apply.
What I do
I write small modules, remote state, and a plan that runs before apply. Naming should make sense to a person who was not in the room. I use the same habit for monitoring config: JSON in, Terraform out, so the review stays readable.
What you get
A module layout and a plan you can read. Not a one-off script that only works on my laptop.

04

Observability and alerting

A few alerts a human can answer, and a dashboard you open during an incident.

The problem
The dashboards exist and nobody trusts the pages. Or there are no dashboards, and every incident starts from a blank terminal.
What I do
I wire metrics to the user path. Prometheus and Grafana, or Dynatrace if that is already what you run. Alerts stay on symptoms. Logs and traces get a first pass where they shorten a diagnosis, and they stay out where they would only add noise.
What you get
A small set of pages and a dashboard that is useful at 2am. Not a wall of charts.

05

Cloud cost and reliability reviews

A short written list of what I would fix first on AWS or Azure, and what I would leave alone.

The problem
The bill moved and nobody is sure which change did it. Or a design will not survive one bad node, one bad zone, or a quiet traffic spike.
What I do
I read the account with you. Waste that is obvious, single paths that take the service down, backup and restore, identity, and the network path. Findings stay in plain language, ordered by what I would touch first.
What you get
A written review. Short. Each item has a reason. I also say what I would not change yet.

06

Ongoing platform support for SaaS startups

Steady help while you do not have a platform team yet.

The problem
You need someone who can keep the platform healthy without hiring a full team this quarter.
What I do
A slice of time, agreed up front. Shipping changes, answering production questions, and improving the path that wasted the most time last month. I work in your repos and your cluster, and I leave notes.
What you get
Help from someone who already knows the system, and a handover if you bring it in house.

How an engagement goes

  1. 01

    Call

    Thirty to forty-five minutes. What hurts, what you already have, and whether I am the right person.

  2. 02

    Audit

    A short look at the cluster, the pipeline, or the cloud account. Written notes, not a slide deck.

  3. 03

    Plan

    Scope, risks, and what I will not touch.

  4. 04

    Build

    Small changes, reviewed, with you in the loop.

  5. 05

    Handover

    Docs, access, and a session so your team can run it without me.

Start a conversation