Projects

Intelligent PDF Query System
LLM / AWS Bedrock / LangChain
Developed an interactive tool for chatting with PDF documents using AWS Bedrock, LangChain, Amazon Titan for embedding generation and similarity search, AWS S3 for storage, AWS EC2 for deployment, and Streamlit for the front-end interface.
View on GitHub
Medical Chatbot: LLM and Vector Embedding-Based Medical QA System
Natural Language Processing / FastAPI / Transformers
Engineered a FastAPI application leveraging transformer-based foundation models i.e., LLMs and Retrieval-Augmented Generation (RAG) in LangChain with SentenceTransformer embeddings and Qdrant vector database, enabling precise medical question-answering and optimized user experience.
View on GitHub
Real-time Object Detection with MobileNet-SSD using OpenCV
Numpy / OpenCV / Python
Deployed the MobileNet-SSD v2 algorithm, trained on the COCO dataset, to detect objects in a live video feed. The project uses OpenCV to implement a pre-trained deep learning model for identifying objects and drawing bounding boxes around them in real-time.
View on GitHub
Artistic Photo Filter Removal Using CNNs
Computational Photography / Image Restoration
Implementation of the paper "Artistic Photo Filter Removal Using CNNs". A method to automatically restore images that have been modified with photographic filters, such as those popular on photo sharing services.
View on GitHub
Facial Recognition using the Viola-Jones Algorithm
Pattern Recognition / Python / Image Processing
Implemented Face-detection using the AdaBoost algorithm, training a face classifier that can distinguish between face and non-face images with improved accuracy.
View on GitHub