Agentic AI Trip Navigator
Agentic AI Trip Navigator is a full-stack AI-powered trip planning application that uses a multi-agent workflow to generate personalized travel itineraries. It includes a modern frontend and a backend system that orchestrates AI agents for intelligent travel recommendations, scheduling, and trip management.
Enterprise Agentic RAG System
Developed a production-ready Enterprise Agentic RAG platform with intelligent multi-agent orchestration, hybrid retrieval (Qdrant + BM25), PostgreSQL integration, and FastAPI APIs for context-aware AI assistance.
Diabetic-Retinopathy-Diagnosis-using-deeplearning
Developed deep learning models to classify the levels of diabetic retinopathy using VGG16, Inception-V3, and a custom CNN with the TensorFlow. Integrated Explainable AI using LIME and Grad-CAM. Built an interactive Streamlit UI and deployed the application on Hugging Face.
Xhalefit MERN Full stack application
Xhale_fit is a full featured personalized fitness coaching web application developed using the MERN stack (MongoDB, Express.js, React, Node.js). The platform empowers users to achieve their fitness goals with custom workout plans, real-time progress tracking, and expert guidance.
AI Doctor_RAG-Based Medical Assistant
AI Doctor is an advanced medical chatbot powered by a Retrieval-Augmented Generation (RAG) system. Leveraging LangChain, OpenAI API, and embeddings, it retrieves relevant medical information and provides accurate, context-aware responses. With a user-friendly Streamlit interface, AI Doctor offers instant, intelligent support for health-related queries.
GEN_AI_Database_QA_Tool
AI-powered question-answering tool designed to interact with relational databases using natural language queries. Built with LangChain, the tool leverages machine learning techniques and prompt engineering to translate user questions into syntactically correct SQL queries, execute them, and return accurate answers.
BERT-Driven-Twitter(X)-Sentiment-Classifier
Twitter (X) sentiment analysis tool using a fine-tuned BERT model. Built with Python, it leverages Hugging Face Transformers for prediction and Streamlit for an interactive UI. Users can input text to get real-time "Positive" or "Negative" sentiment predictions with confidence scores.
Gen AI-Powered Log Classification with Regex, Clustering, and Regression
Hybrid Log Classification System designed to intelligently analyze and classify system logs. The system integrates LLM, Regex-based preprocessing, Sentence Transformers, DBSCAN clustering, and Logistic Regression into a unified pipeline.
Agentic AI-Based E-commerce Price Tracker and Smart Buyer
An Agentic AI system that acts as a smart e-commerce assistant, interpreting user-defined purchase rules through natural language and autonomously evaluating conditions to simulate buying actions. It uses OpenAI's LLM for rule extraction, Selenium for live data scraping, and Streamlit for the UI.
React_Based_Movie Searcher_App
A frontend web app built with React that lets users search for movies and save favorites. It integrates a public movie API, uses Context API for state management, and features a responsive UI with a favorites page for a personalized experience.
NEWS_Articles_Q&A_Tool_With_Langchain
This project is a Generative AI (GenAI) tool designed for automated question answering on news articles. It uses NLP techniques with LangChain and OpenAI API to extract and generate contextually accurate answers. The UI is built using Streamlit, enabling users to input articles and ask questions.
house price prediction using machine learning
I developed a machine learning model to predict house prices based on various features such as location, size, and amenities. The end-to-end solution includes data preprocessing, feature engineering, model training, and evaluation using algorithms like Linear Regression, Random Forest, and XGBoost.
Potato_Plant_Health_Classifier_Using_deeplearning
This project is a deep learning-based tool for classifying the health of potato plants. Utilizing Convolutional Neural Networks (CNNs) and built with TensorFlow, the classifier can accurately identify whether a potato plant is healthy or affected by common diseases.
Laptops Online Store-Website Built Using HTML, CSS, and JavaScript
I developed a fully responsive laptop e-commerce website using HTML, CSS, and JavaScript. The platform allows users to browse laptops, view product details, add items to their cart and filter products based on various criteria such as price, brand, and specifications. The site provides a seamless and intuitive user experience.
