RAG Chat Agent
Designed and built a local RAG-based chat agent to retrieve, analyze, and summarize official documents using LLM-powered reasoning. Implemented real-time document retrieval with tool-assisted responses and an asynchronous FastAPI backend. Technologies: Python, FastAPI, MySQL, Ollama (Qwen2), AsyncIO, HTML/CSS, JavaScript.
Recommendation System
Developed a product/content recommendation engine using collaborative filtering techniques. Used LightFM and Scikit-learn to model user-item interactions and generate personalized suggestions. Technologies: Python, TensorFlow, LightFM, Scikit-learn, Pandas, Matplotlib.
Sales Performance Analysis
Analyzed large sales datasets to surface trends, top-performing products, and regional performance gaps. Created visual dashboards and automated reports to support data-driven business decision-making. Technologies: Python, SQL, Pandas, Matplotlib, Seaborn, Excel.
Object Detection Pipeline
Implemented an object detection system to identify and label objects in images using bounding boxes. Improved model detection accuracy through iterative tuning and result visualization with OpenCV. Technologies: Python, OpenCV, TensorFlow, Matplotlib.