Smart News Digest Engine

This project is an LLM-powered news summarization tool that retrieves and condenses real-time news articles based on user-provided topics. Built using Python and llama 3.2 LLM, it combines a Streamlit UI, autonomous agents, and natural language processing.

Stress Analysis using Natural Language Processing (NLP)

I built an NLP based stress analysis system using statistical analysis, Vader and Pattern sentiment analysis, RoBERTa model and transformer pipeline to quantify stress and mental health status of Reddit's users. Visualized and compared how each models compare against each other.

PCA from scratch without using scikit-learn library

This is a project which explores Principal Component Analysis in data science but without using any ML library like scikit-learn. It is a scratch implementation. I also worked to restore the original data. This is a scratch step-by-step implementation.

TheGaucStop - An end-to-end cloud data warehouse and analytics solution

Implemented a complete cloud data warehouse solution designed to store, manage, and analyze business data. Developed the workflow using Oracle Cloud Infrastructure, Apache Hop, and ETL processes automating the extraction, transformation, and loading of data into both dimension table and fact table.

AI Resume Assistant

I built a Streamlit app that scores resumes vs job descriptions (TF-IDF + cosine similarity), detects missing skills, and uses GPT-3.5 to enhance bullets, generate tailored cover letters, and deliver smart feedback reports. The project is built using Python, scikit-learn, NLTK and Streamlit.