About Me
Learn more about my background and experience.
Highly motivated Software Engineer with a Master’s in Computer Science (GPA 4.0) from the University of Colorado Boulder, skilled in full-stack development, backend systems, and high-performance C++ programming. Experienced in building scalable web applications, AI agentic servers, and optimized data structures like LRU caches, with strong expertise in Python, C++, Java, JavaScript/TypeScript, SQL/NoSQL, and cloud deployment (AWS). Demonstrated ability to improve system performance, implement multithreaded and read-optimized solutions, and deliver robust, user-focused software in Agile environments. I’m also passionate about using AI to solve meaningful problems and enjoy working on teams that value innovation, performance, and social impact.
Skills & Technologies
A comprehensive overview of my technical skills and the technologies I work with.
Work Experience
My professional journey and key accomplishments.
Created Backend APIs for MCP Servers
Developed Website and an App called Flow State
Education
My academic journey and educational background.
University of Colorado Boulder
Specialized in Artificial Intelligence and Machine Learning
Certifications
Professional certifications and achievements that validate my expertise.
Featured Projects
A showcase of my recent work and personal projects that demonstrate my skills and creativity.
Flow State App
Flow State App
Designed and deployed a full-stack productivity app combining Pomodoro timer, meditation, and task-linked focus cycles, tailored for displaced students in Myanmar IDP Refugee camps. Used static generation and cloud deployment with CI/CD via GitHub and Vercel. Supported 100+ users; aimed to enhance focus, engagement, and mental health using behavioral reinforcement loops.
Personal Portfolio Website
Personal Portfolio Website
Built and deployed a fully static, SEO-optimized developer portfolio site with custom project showcase and continuous deployment.
Microservices — Movie Rating Application
Microservices — Movie Rating Application
This project showcases a microservices-based backend architecture for a simple movie-rating application similar to IMDB. Each microservice runs independently and communicates via REST using Eureka for service discovery.
Cinema Ticket Management System
Cinema Ticket Management System
Movie Booking System built with ASP.NET Core MVC and MS SQL Server.About Movie Booking System built with ASP.NET Core MVC and MS SQL Server. It manages movie screenings and ticket reservations, offering distinct features for Members, Content Admins, and System Admins.
Tiny LRU Cache
Tiny LRU Cache
About LRU Cache implemented using Object Oriented Principles (C++)
Parking Management System
Parking Management System
The Parking Management System is a multi-floor parking management application designed to facilitate vehicle check-in and check-out, live occupancy tracking, role-based access, flexible slot assignment, and payment integration. This system aims to streamline parking operations and enhance user experience.
MCP_Server
MCP_Server
This project demonstrates a multi-container Model Context Protocol (MCP) system consisting of two independent microservices: Golang MCP Server → provides Mathematical Tooling; Python MCP Server → provides Web Search Tooling; Python Client → interacts with both servers through Docker’s internal network
Customer Segmentation with Unsupervised Learning
Customer Segmentation with Unsupervised Learning
Segmented customers in an online retail dataset using K-Means and Hierarchical Clustering; applied EDA, outlier detection, transformation, and standardization; tuned hyperparameters using Elbow and Silhouette methods to identify meaningful clusters for business insights. Engineered features and removed outliers to boost cluster interpretability by 50%
Lung Cancer Detection (CT Image-Based)
Lung Cancer Detection (CT Image-Based)
Trained multiple supervised models on lung CT scan data to assist physicians in early cancer detection. Achieved 99% accuracy with SVM; documented full model evaluation in published report. Compared models using ROC and precision-recall; aimed to reduce clinical turnaround time
BBC News Classifier with Supervised and Unsupervised Models
BBC News Classifier with Supervised and Unsupervised Models
Built a news classification system trained on the BBC dataset using both supervised (SVC: 98.5% accuracy) and unsupervised (NMF: 95.5% accuracy) methods. Preprocessed and vectorized news articles to categorize into business, politics, tech, and sports.
Retrieval-Augmented Generation for AWS Exam Q&A
Retrieval-Augmented Generation for AWS Exam Q&A
Built a RAG-based system for AWS certification exam questions that eliminated the need for costly fine-tuning on extensive documentation. By creating a dataset with LLMs from presentation slides, the system avoided retraining with documentation updates. This solution improved performance by 24%, achieving a higher accuracy than fine-tuned Gemma3 models.
Get In Touch
I'm always open to discussing new opportunities, interesting projects, or just having a chat about technology.