Anurag Dhami

A Python Flask application that uses Google's Gemini AI to automatically parse Statement of Facts (SoF) documents. It accepts PDF and Word files, extracts all port operation events with their start/end times, and provides the structured data for download as JSON or CSV.

✍️ An interactive AI application that solves math problems from hand gestures. Draw equations in the air using your webcam and get instant solutions from Google's Gemini AI. Built with Python, Streamlit, and MediaPipe.

📈 This project forecasts retail demand by comparing 3 time series models: Holt-Winters, Auto-ARIMAX, and Facebook Prophet. It performs a complete data science workflow, including EDA and feature engineering with external regressors to predict future sales from historical data. Model performance is evaluated using RMSE, MAE, and MAPE.

ChatScholar is an AI-powered web application that leverages RAG (Retrieval-Augmented Generation) for intelligent PDF Q&A and automated essay grading using Google's Gemini model. It helps students, educators, and professionals quickly extract knowledge from documents and receive detailed writing feedback.

MediClaim is an AI-powered web application designed to streamline and automate the health insurance claim verification process. It leverages a Retrieval-Augmented Generation (RAG) pipeline with Google's Gemini large language model to analyze submitted claims against a knowledge base of insurance policy documents.

ApnaGhar is a MERN stack-based web application designed to help college students easily find rooms for rent near their campuses.

important question with solution