AI-powered YouTube comment analytics platform using NLP and Transformers to extract sentiment, audience insights, and interactive RAG chat.

Credit card fraud detection and anomaly detection pipeline using PCA, t-SNE, K-Means, DBSCAN, Isolation Forest, and One-Class SVM on imbalanced transactional data.

Customer segmentation project using K-Means, DBSCAN, and PCA/t-SNE to analyze supermarket spending habits and optimize targeted marketing strategies.

Real-time endpoint log anomaly detection system using Isolation Forest, K-Means clustering, and graph analysis for Windows Security Event Logs monitoring and threat detection.

Prototype version of SentimentPRO (Streamlit + Python) for YouTube comment sentiment analysis, featuring NLP models (MARBERT/RoBERTa), analytics dashboards, and SQL Server storage.