RANDHIR KUMAR

Text2SQL-Streamlit is a Natural Language to SQL (NL2SQL) search application that allows users to query a PostgreSQL database using plain English. The system translates user queries into SQL, executes them safely, and displays results through an interactive Streamlit interface. It also supports hybrid search using pgvector for semantic similarity.

This project simulates real-time video stream processing to detect people and trigger crowd alerts using the YOLOv5 object detection model.

A Streamlit-based AI chatbot designed to answer questions from PDF documents by combining semantic search and large language model generation.

This project is a Retrieval-Augmented Generation (RAG) based chatbot that lets you upload PDF or TXT documents and ask questions about their content — all running locally (⚡ no API keys or internet dependency required after setup).

Config files for my GitHub profile.

AI-powered YouTube Q&A Assistant – Download, transcribe, search, and summarize video content using Whisper, FAISS, and Streamlit.

This Python script uses OpenCV and a pre-trained Keras deep learning model to perform real-time gender detection through a webcam.

A Python + MySQL desktop application to manage LPG (Liquefied Petroleum Gas) services. It provides an easy-to-use interface for users to register, request connections, and book cylinders, while allowing the admin to manage users, track bookings, generate reports, and modify system data.

The code sets up a Flask web application for a chatbot using a pre-trained model. It loads intents from a JSON file, tokenizes patterns with CountVectorizer, and defines routes for user input, predicting tags, and retrieving bot responses. The application runs on localhost when the script is executed.