Stop Saying "It Works on My Machine": Containerizing a B2B SaaS with Docker

RIDHAM POKIYA — Fri, 29 May 2026 10:02:20 +0000

When you are a two-person startup building an enterprise SaaS, speed is everything. My brother, Tushar, handles the MERN stack development for Heeratrack, while I manage the DevOps and infrastructure.

In the early days, we hit the classic developer bottleneck: “It works on my machine, why is it crashing on yours?” Differences in Node versions, missing environment variables, and local database mismatches were slowing us down.

To fix this, I completely containerized our workflow using Docker.

How We Use Docker for Heeratrack:

The Multi-Container Setup
I wrote a docker-compose.yml file that spins up our entire stack locally in seconds. With one command, we boot up the React frontend, the Node/Express backend, and our database. No manual installations required.
Eradicating Environment Mismatches
By defining strict Dockerfiles for both the frontend and backend, we ensure that the exact same environment used for local development is the exact same environment deployed to production. This gives us 100% confidence when we push new code.
Seamless CI/CD Integration
Containerizing the app was step one. Step two was plugging it into our GitHub Actions pipeline. Now, whenever Tushar merges a pull request, our pipeline automatically builds the Docker images, tests them, and pushes them to our container registry.

For any DevOps engineers working with frontend developers: do not wait to implement Docker. Containerize on Day 1. It saves you hundreds of hours of debugging later.

Architecting a SaaS from Scratch: The AWS Infrastructure Behind Heeratrack

RIDHAM POKIYA — Thu, 28 May 2026 06:48:26 +0000

Over the past few months, I’ve been documenting my technical journey through the #100DaysOfDevOps challenge, diving deep into tools like Terraform, Kubernetes, and AWS Application Load Balancers. Today, I want to share exactly what I am building all this infrastructure for.

I am currently building Heeratrack, a B2B SaaS platform designed to modernize inventory management and worker allocation for the diamond manufacturing industry in Surat, India.

The Architecture Challenge
The diamond industry operates on micro-milligrams and strict security. A dropped database connection or a leaked inventory sheet is unacceptable. Here is how I am approaching the infrastructure:

High Availability with AWS EC2 & Auto Scaling
We cannot afford downtime during peak trading hours. I am deploying the core backend using Application Load Balancers (ALB) routing traffic to auto-scaling groups of EC2 instances. If traffic spikes, the infrastructure scales horizontally without manual intervention.
Infrastructure as Code (IaC)
To ensure our environments (Development, Staging, Production) are identical and easily reproducible, I am writing the entire infrastructure using Terraform. This eliminates configuration drift and allows us to tear down or rebuild environments in minutes.
Automated CI/CD Pipelines
As a two-person team, we need to move fast. I am setting up automated pipelines (using GitHub Actions and Jenkins) so that when my brother pushes a new feature in the MERN stack, it is automatically tested, containerized via Docker, and deployed seamlessly.

Building the infrastructure for a real-world SaaS is the ultimate test of my DevOps skills. In my upcoming posts, I will be breaking down the exact Terraform scripts, Kubernetes manifests, and AWS configurations I am using to build Heeratrack.

Let me know in the comments if you want a deep dive into any specific part of the architecture!
connect with me:
https://github.com/ridhampokiya2110
https://www.linkedin.com/in/ridham-pokiya-b7974a249
https://ridhampokiya.netlify.app/

DEV Community: RIDHAM POKIYA

Stop Saying "It Works on My Machine": Containerizing a B2B SaaS with Docker

Architecting a SaaS from Scratch: The AWS Infrastructure Behind Heeratrack