YouTube AI Analyzer is a full-stack application that evaluates YouTube videos and comments using AI. It offers summaries, sentiment analysis, and interactive visualizations for efficient video comparison and feedback extraction.

Portfolio AI Assistant is an agentic Retrieval-Augmented Generation (RAG) chatbot embedded in my portfolio. It provides precise answers about my projects, skills, certifications, and learning journey, handling multi-turn conversations with context-aware retrieval and hybrid search.

The IELTS Essay Evaluator is an AI-powered application designed to assess IELTS Writing Task 2 essays based on official IELTS scoring criteria

A fine-tuned multilingual XLM-RoBERTa model optimized for YouTube comment sentiment analysis. Trained on over 1 million labeled comments, it significantly improves accuracy in detecting positive, neutral, and negative sentiments compared to the Twitter-trained baseline.

A machine learning model for detecting fraudulent financial transactions using classification models, feature engineering, and SMOTE to handle class imbalance.

A neural network model that predicts whether a customer will make another audiobook purchase based on past interactions and engagement data.

Curated a large-scale dataset of 1M+ YouTube comments annotated with sentiment labels (Positive, Neutral, Negative) across multiple domains and countries, enriched with metadata for NLP, sentiment analysis, and LLM fine-tuning.

A data curation & analysis project on IMDb’s Top 250 movies—scraped using Scrapy and analyzed using Python to uncover trends in genres, ratings, and movie durations through rich visualizations.

A Selenium-based script that automates Twitter login, scrapes tweets from the feed, and stores them in a CSV file with duplicate prevention and dynamic scrolling support.

A real-time multi-tenant chat system with strict tenant isolation, role-based access control, and secure authentication. Designed for organizations that need private communication environments with Super Admin oversight and Admin-managed tenants.

Built a dynamic application using vanilla JavaScript and SCSS that allows users to search over a million recipes, view detailed instructions, and upload their own recipes. Integrated with the Forkify API.

Developed a responsive music player web app using **Next.js**, providing users with an intuitive interface to browse and listen to audio tracks seamlessly.