This project enables LLMs to process handwritten documents by converting them into machine-readable text using CNNs, RNNs, and Scale Space Technique, achieving 91% accuracy. It integrates LLaMA2, Hugging Face, and FAISS for real-time NLP interaction, making handwritten text searchable for research, education, and automation.

This project presents a CNN-based smoke detection model for laparoscopic surgeries, improving visibility in electrocauterization procedures. Using Normalized-RGB, Wavelet energy, GLCM, HSV, Optical Flow, and Fog Area, it achieves 92.6% accuracy on annotated robot-assisted surgery videos, enhancing real-time surgical assistance and reducing risks.

Pioneered an innovative ensemble learning approach to predict movie success, achieving 94.6% accuracy by integrating diverse data sources. Utilized stacking ensemble algorithms, including Logistic Regression, KNN, AdaBoost, Decision Tree, XGBoost, and Random Forest.

The Smart Campus ID System is a cutting-edge solution designed to enhance student identification, attendance tracking, and campus security using NFC, GPS, and mobile payment systems. This all-in-one platform streamlines administrative tasks and improves student safety while ensuring seamless operations. Also filed a patent for the project.

KisanMadad is a platform that uses AI, IoT, and drones to connect farmers directly with markets, optimize irrigation, monitor crop health, and provide educational resources, enhancing productivity and profitability.

This project leverages MediaPipe and LSTM-based neural networks to develop a real-time sign language interpretation system. By capturing keypoint detection and temporal dynamics of gestures, it promotes seamless communication and inclusivity for the hearing-impaired community.

Introducing our custom Python chatbot, powered by cutting-edge libraries like Langchain, Llama2, Faiss, Hugging Chain, and Gradio. Simply provide a document, and it effortlessly responds to all your document-related questions with accuracy and speed.

VTOP Mobile Application is a Java-based student management portal for VIT Chennai College, utilizing Firebase for secure data storage. It enables leave application submission, proctor messaging, credential retrieval, and offers a comprehensive FAQ section, enhancing the student experience and streamlining administrative processes.

Developed a cutting-edge prototype web application for VIT Chennai College's Placement Cell .VIPAT offers a seamless and efficient digital solution that replaces the manual process.

BigO is a web application that offers an AI-powered chatbot for conversational interactions. It utilizes the OpenAI API to provide intelligent and informative responses to user queries.