Projects 💻

Technologies: Go, gRPC, WebSocket, AWS, Docker, Protocol Buffers, Next.js

• Developed a real-time document platform using Go and WebSocket communication for live updates and user activity synchronization.
• Enhanced user experience by integrating gRPC and Protocol Buffers for efficient communication and DynamoDB for data storage.
• Utilized AWS EC2 for scalable deployment and Docker for containerization, ensuring high availability and portability of the service.

Technologies: Go, gRPC, WebSocket, AWS, Docker, Protocol Buffers, Next.js

• Constructed a Streamlit-based Question-Answering application integrating LangChain and OpenAI’s GPT-3.5 Turbo model for efficient natural language processing of PDF, DOCX, and TXT files.
• Implemented features include chunking, embedding creation using OpenAIEmbeddings, and persistent chat history, showcasing proficiency in document processing, vector storage, and user interface design.

Technologies: React, TypeScript, Yjs, WebRTC, Monaco Editor

• Programmed a collaborative code editor using React, Yjs, WebRTC, & Monaco Editor, enabling real-time communication.
• Enabled real-time code synchronization, shared output, and persistent peer sessions to support seamless multi-user collaboration.

Technologies: Go, Gin, REST API, HTML, JavaScript, Render, GitHub Pages

• Built a high-performance URL shortening service with Go and Gin, featuring real-time analytics and RESTful API endpoints.
• Designed a modern, interactive demo page for users to shorten links and view analytics live.
• Implemented robust error handling, input validation, and a hardcoded demo short link for portfolio reliability.

Technologies: C++, Process Scheduling, Memory Management, OS Simulation

• Simulated an operating system that handles multiple tasks using round-robin scheduling for fair CPU time allocation.
• Implemented paging to manage memory efficiently and support multitasking on limited resources.
• Added interrupt handling and error recovery to maintain stability during high-load and faulty execution scenarios.

Technologies: Python, TensorFlow, ResNet18, OpenCV, Medical Imaging, AI

• Built a deep learning model using ResNet18 to detect retinal diseases like diabetic retinopathy and glaucoma with 97% accuracy.
• Applied advanced image preprocessing (contrast enhancement, denoising, augmentation) to clean low-quality scans and improve AI reliability.
• Trained the model to identify subtle changes in blood vessels, enabling faster and more consistent diagnosis with less manual review.
• Fine-tuned the pipeline for real-world usage across hospital systems by testing it on diverse datasets and imaging devices.

Technologies: Python, TensorFlow, SSD Object Detection API, OpenCV, Computer Vision

• Built a real-time sign language recognition system using TensorFlow and the SSD object detection framework, achieving 95% accuracy.
• Trained on a custom dataset of 20,000+ gesture images, using data augmentation to improve model performance and robustness.
• Integrated OpenCV for live video capture, enabling instant translation of hand gestures into readable text in real time.

Technologies: Python, Java, AST, Compiler Design, Multithreading

• Built a tool to convert C code into Java and Python using ASTs, reducing manual translation effort and errors.
• Handled complex C constructs (pointers, macros, memory management) by applying compiler design strategies.
• Optimized conversion speed to 1,000+ lines/min using multithreading, making it suitable for large legacy projects.