Cloud, Machine Learning & Robotics Projects

Zero-Trust Resume Tailor Platform on AWS

Designed and provisioned AWS infrastructure with Terraform for a production grade web application deployment.

AWS | Terraform | Zero Trust
  • Automated instance bootstrap with cloud-init and systemd to install dependencies, configure the Python backend, and manage application startup and recovery on Ubuntu EC2.
  • Integrated AWS SSM Parameter Store for secure retrieval of the OpenAI API key and Cloudflare tunnel token, eliminating hardcoded secrets from infrastructure and runtime configuration.
  • Replaced direct public exposure with Cloudflare Tunnel and Access, routing traffic and enforcing email-based Zero Trust authentication for controlled access.

End-to-End DevOps & GitOps Implementation for Microservices E-Commerce Platform

Designed and implemented a microservices-based DevOps architecture for an e-commerce application using Docker, Kubernetes, and AWS EKS.

DevOps | Kubernetes | EKS
  • Provisioned cloud infrastructure using Terraform including VPC, IAM, and EKS with remote state management, backend configuration, and state locking.
  • Containerized services with Docker, orchestrated local development using Docker Compose, and deployed workloads to Kubernetes clusters.
  • Implemented Kubernetes Deployments, Services, and Ingress resources with an Ingress Controller for external traffic routing.
  • Automated application delivery using GitOps with Argo CD, enabling declarative and version-controlled Kubernetes deployments.
  • Built CI pipelines using GitHub Actions to build and push container images and integrated them with Argo CD for end-to-end CI/CD.

Smart Receipt Scanner - Cloud-Native Serverless Project

Built a scalable serverless receipt processing system on AWS using Terraform, Lambda, S3, API Gateway, and DynamoDB to support multi-user workloads.

Serverless | OCR | AWS
  • Implemented secure authentication and access control with AWS Cognito and IAM roles, ensuring strict data isolation.
  • Developed real-time OCR pipelines in Python using Tesseract, triggered by S3 events to extract structured receipt data for analytics.

One Shot GenAI Bot on AWS using Bedrock

Designed and provisioned AWS infrastructure using Terraform to deploy EC2 instances, security groups, and IAM roles following least privilege principles.

GenAI | Bedrock | AWS
  • Integrated Amazon Bedrock with Python and Boto3 to invoke Amazon Nova 2, with configurable support for Anthropic Claude Sonnet 4.5 via the Bedrock Runtime API.
  • Built and deployed a one-shot text processing web application using Apache and Python CGI to deliver single-response LLM inference.

Deployment of a Scalable and Secure WordPress CMS on AWS

Built a highly available and scalable WordPress architecture using multi-AZ EC2, Auto Scaling, and ALB with LAMP and LEMP stack configurations.

AWS | WordPress | HA
  • Performed MariaDB backup and restoration on Amazon RDS using SQL dumps to support environment migration.
  • Reduced infrastructure costs by approximately 21% through Terraform automation, VPC endpoints for S3, SSM, and KMS, and optimized network design.
  • Enhanced security and reliability using VPC segmentation, IAM roles, security groups, CloudWatch monitoring, and SNS alerts.

AWS Serverless Data Processing Pipeline

Developed an event-driven pipeline that processes text files uploaded to Amazon S3, performs word frequency analysis, and stores results in DynamoDB.

AWS | Step Functions | Lambda
  • Orchestrated multiple Lambda functions using AWS Step Functions to enable fault-tolerant execution and real-time SNS notifications.
  • Achieved scalable and cost-efficient processing through serverless architecture with AWS auto-scaling and pay-per-use pricing.
  • Implemented CloudWatch monitoring and IAM-based security controls to ensure performance visibility, reliability, and compliance with AWS best practices.

English Accent Classifier (Streamlit)

Developed and published a Streamlit web application for classifying English accents from YouTube video links.

Deep Learning | Web App
  • Utilized a pretrained Speech Brain model for robust audio classification.
  • Implemented audio extraction, preprocessing, and real-time inference pipelines with PyTorch and torch audio.
Completed: Jun 2025 View Live Demo

End-to-End ML Pipeline with DVC, MLflow & AWS

Created a fully reproducible ML workflow for model versioning, experiment tracking, and deployment.

MLOps | Data Versioning
  • Used DVC for versioning and MLflow for tracking experiments (Random Forest on medical data).
  • Leveraged Dags Hub for remote storage and deployed the MLflow server on AWS EC2.
  • Containerized the entire workflow using Docker for scalability.
Completed: Mar-Apr 2025 Private Repository

Scribble-based Image Segmentation with Constraints

Designed a deep neural network leveraging minimal scribble-based user input and convexity constraints for robust region delineation.

Deep Learning | Computer Vision
Completed: Dec 2023 View Code

Pick and Place Robotic Arm (LEGO EV3)

Developed an inverse kinematics model for a 3 DOF manipulator robot to execute precise pick, place, and homing operations.

Robotics | Control Systems
  • Implemented operations using PID controllers and feedback from encoders/sensors.
  • Programmed the system entirely in MATLAB.
Completed: Mar 2022

LLMs Implementation Workshop

Hands-on mastery of foundational LLM concepts, including tokenizer development, architecture coding, and fine-tuning models like GPT-2 and Llama 2.

LLMs | Deep Learning
Completed: Jan 2025

Additional Projects

Selected deep learning, robustness, and systems work grouped into a single expandable project card.

Deep Learning Classification

Trained a supervised CNN (TensorFlow Keras) on a Simpsons character dataset (90% accuracy) and used Transfer Learning (PyTorch) on a custom dataset (89% accuracy).

Data Augmentation (AugMix)

Integrated ResNet-18 and ConvNeXt-Tiny into the AugMix framework to evaluate model robustness under distribution shifts, tracking results with TensorBoard. Code

Image Column Sorter (C++/Python)

Modified a CLI tool to sort BMP image columns by average RGB value using C++17 and replicated the logic in Python. Used CMake for builds.