DevOps Portfolio

v4.0.2-stable

Portfolio

Project

Scalable Knowledge-Driven RAG API

A scalable RAG API which answers questions by searching a knowledge base, then using AI to generate accurate answers

Project

E2E Web App Automation on AWS

Deployed a containerized app on AWS EKS with Terraform and automated CI/CD pipelines for zero-touch deployments.

Project

React App Deployment with Automated CI/CD Pipeline

Deployed a Dockerized React app on AWS with CI/CD, GitOps workflows, and monitoring.

Project

Web App deployment using AWS CodeBuild/CodePipeline

Containerized and deployed a Web app on AWS with CI/CD using CodeBuild and CodePipeline

analytics

RAG-Powered API Architecture

Project Deep Dive

Case Study 01

$ ./project --launch

I spent my afternoon building a production-level API with RAG (Retrieval-Augmented Generation). The goal was to move beyond simple chatbots and create a robust middleware that can extract semantic meaning from unstructured data and pipe it into localized LLMs for precision responses.

code

git push origin master

VIEW SOURCE ON GITHUB

arrow_forward

systemctl status

BUILD_STATUS:SUCCESS

RUNTIME_LATENCY:120ms

MEMORY_USAGE:256.4 MB

LAST_DEPLOY:2026.05.24_14:30

Services & Concepts

database

ChromaDB

Vector storage for fast semantic search and retrieval.

smart_toy

Ollama

Running local LLMs for data privacy and low-latency.

api

FastAPI

High-concurrency endpoints for production-scale usage.

hub

Kubernetes

Container orchestration for production-scale usage.

code

Python

Developing the API code for the RAG system.

commit

Github Actions

Gitops workflow for building and deploying the dockerized image

Challenges & Wins

The primary challenge was designing an efficient retrieval pipeline, ensuring relevant context selection, and optimizing latency while integrating vector search with FastAPI. But in the end built a scalable, containerized RAG API delivering accurate, context-aware responses, with production-ready deployment on Kubernetes and automated CI/CD pipelines.

verified Generating accurate LLM answers

Why I did this project

Motivation stemmed from a desire to gain hands-on knowledge in the synergy between FastAPI and RAG. I wanted to demonstrate that AI isn't just a wrapper but a core architectural component.

school Growth Focused

Ready to explore the code?

The entire repository is public and documented.

code View Source on GitHub arrow_forward

Terminal Pulse v1.0.4 • Project Deployment Report

cloud

AWS Powered Deployment Architecture

Project Deep Dive

Case Study 02

$ ./project --launch

Designed and deployed a production-grade containerized application on AWS EKS using Docker, Kubernetes, and Terraform. The project focuses on building a fully automated DevOps pipeline with Jenkins and GitHub Webhooks, enabling zero-touch deployments, infrastructure as code provisioning, and scalable cloud-native architecture.

code

git push origin master

VIEW SOURCE ON GITHUB

arrow_forward

systemctl status

BUILD_STATUS:SUCCESS

RUNTIME_LATENCY:120ms

MEMORY_USAGE:256.4 MB

LAST_DEPLOY:2026.05.24_14:30

Services & Concepts

code

Terraform

Infrastructure as Code for provisioning and managing cloud resources.

hub

Kubernetes

Container orchestration for scalable and resilient deployments.

commit

Jenkins

Automated Declarative CI/CD pipelines for build, test, and deployment.

analytics

Uptime Kuma

Leveraged open source tool for monitoring and App alerting during downtime

cases

Docker

Containerization for packaging and deploying applications

Challenges & Wins

Faced Terraform provisioning failures including duplicate Key Pair conflicts, missing IAM EKS permissions, alongside Jenkins pipeline failures due to Java 11/17 mismatch, missing kubectl on EC2, and sudo permission errors. Resolved EKS authentication failures caused by outdated AWS CLI generating deprecated v1alpha1 API, fixed by upgrading to AWS CLI v2, updating aws-auth ConfigMap, and copying kubeconfig to Jenkins user. Integrated Uptime Kuma for real-time application health monitoring, achieving a fully automated zero-touch deployment pipeline.

Overcame all blockers through systematic debugging, delivering a production-grade Cloud-Native CI/CD setup on AWS.

Handyman Configuration Management

Why I did this project

I wanted to move beyond manual cloud setups and get real hands-on experience with Terraform — understanding how infrastructure is defined as code and provisioned automatically.

school Growth Focused

Ready to explore the code?

The entire repository is public and documented.

code View Source on GitHub arrow_forward

Terminal Pulse v1.0.4 • Project Deployment Report

folder

React App deployment on AWS

Project Deep Dive

Case Study 03

$ ./project --launch

Containerized a React application using Docker & Nginx, built a fully automated Jenkins CI/CD pipeline triggered by GitHub Webhooks. Dev pushes auto-deploy to a public repo and prod merges promote to a private repo on DockerHub. Deployed and secured the application on AWS EC2 with strict network policies, automated the build & deploy lifecycle via Bash scripts, and set up Uptime Kuma for 24/7 health monitoring with instant downtime alerts.

code

git push origin master

VIEW SOURCE ON GITHUB

arrow_forward

systemctl status

BUILD_STATUS:SUCCESS

RUNTIME_LATENCY:60ms

MEMORY_USAGE:128.4 MB

LAST_DEPLOY:2026.05.24_14:30

Services & Concepts

cloud

AWS

Provisioning and managing cloud resources.

commit

Jenkins

Automated Declarative CI/CD pipelines for build, test, and deployment.

cases

Docker

Containerization for packaging and versioning images to DockerHub

Challenges & Wins

Started strong but hit some real walls, Jenkins on Windows fought back hard with shell compatibility issues, Git Bash mangling Unix paths, and ngrok URLs breaking webhook triggers every restart. GitHub Pages also failed to serve the React app correctly due to asset path conflicts, forcing a switch to Netlify. Pushed through all of it and came out with a fully automated CI/CD pipeline where a single git push handles everything - build, push, deploy - across dev and prod environments.

Topped it off with AWS EC2, Nginx, and Uptime Kuma running 24/7 health checks, making it a genuinely production-ready setup.

Compare Environment compatibility

Why I did this project

Built this as a side project to get my hands dirty with DevOps, wanted to see experiment with all the pieces fit together in a real deployment workflow rather than just reading about it.

Science Experimentation

Ready to explore the code?

The entire repository is public and documented.

code View Source on GitHub arrow_forward

Terminal Pulse v1.0.4 • Project Deployment Report

analytics

CI/CD Pipeline with CodeBuild, ECR & EKS on AWS

Project Deep Dive

Case Study 04

$ ./project --launch

I took a simple static HTML app and containerized it using Docker with Nginx, pushed the image to AWS ECR, deployed it on a Kubernetes cluster running on AWS EKS, and automated the entire build and deploy process using AWS CodeBuild and CodePipeline. So every time we push code to GitHub, it automatically builds, pushes to ECR, and deploys to Kubernetes.

code

git push origin master

VIEW SOURCE ON GITHUB

arrow_forward

systemctl status

BUILD_STATUS:SUCCESS

RUNTIME_LATENCY:40ms

MEMORY_USAGE:456.4 MB

LAST_DEPLOY:2026.05.24_14:30

Services & Concepts

hub

AWS EKS

Managed Kubernetes cluster for deploying and managing containerized applications.

commit

CodeBuild/CodePipeline

AWS resources for automating the build and deploy process.

code

YAML

Scripting the Configuration for pipelines and Kubernetes deployment and services

Challenges & Wins

Ran into a pod scheduling issue on EKS where the nodes were too full to run our app pods, which I quickly fixed by reducing the replicas to 1 and also selected t3.micro instance type for the nodes due to resource constraints with AWS account. But despite those bumps, got a fully automated CI/CD pipeline running from a simple HTML file all the way to a live Kubernetes deployment on AWS with just a git push

event_busy Resource constraints

Why I did this project

Motivation stemmed to explore AWS CodeBuild and CodePipeline. So grabbed a random app, threw it into Docker, pushed it to EKS, and wired up a full CI/CD pipeline just to see how the AWS pieces fitting together.

explore Trying out new things

Ready to explore the code?

The entire repository is public and documented.

code View Source on GitHub arrow_forward

Terminal Pulse v1.0.4 • Project Deployment Report