Projects
My collection of professional and personal projects that I have worked on from internships to personal projects.
DyMoN: A Post-Hoc Kinematic Approach for Detecting Anomalies in Dynamic Networks
A from-scratch kinematic model for anomaly detection in dynamic graphs. It tracks network evolution through physical motion metrics, calculating magnitude and direction scores—specifically isolating the tanh function for directionality. The model was benchmarked against DSEDN, JODIE, and STRIPE baselines to validate its motion-aware approach.
Personal Portfolio with CMS
A full-stack Next.js portfolio and CMS self-hosted on Proxmox via Docker, featuring traffic analytics and GitHub OAuth security. Self-hosted with CI/CD via Github Runners and utilizes Tailscale Funnel for secure, global public access.
Real-Time Edge LPR Pipeline
A real-time License Plate Recognition (LPR) system using fine-tuned YOLO and PaddleOCR models, optimized with OpenVINO for high-performance execution on edge hardware, published via MQTT, and containerized using Docker for resource-constrained deployment.
The Homelab
A "trash-to-treasure" infrastructure project resuscitating a legacy i5 laptop into a Proxmox-powered private cloud. It orchestrates a Dockerized CI/CD pipeline for my portfolio while managing network-level security and real-time visitor analytics.
BEMMRA: An Application for Bus and Employee Management for Maintenance and Route Assignment
An offline, centralized database system designed to manage HR, maintenance, and route scheduling for a large-scale bus company. Built with role-based access control to eliminate data redundancy across multiple base departments.
AdaptiveDo
A User Modelling Proof-of-Concept, AdaptiveDo is a responsive task management application built with Next.js and TypeScript. It demonstrates adaptive interface design by automatically adjusting the UI based on user workload, overdue tasks, and user experience level.