Projects
Welcome to my portfolio! This is where I showcase some of the exciting projects I've worked on. Feel free to take a look around and explore the innovative solutions I've contributed to.
Project 01: FraudDetectAI
FraudDetectAI is a sophisticated NLP system designed to analyze chat messages and identify fraudulent behavior. The solution leverages BERT, a cutting-edge transformer-based model, to extract meaningful features from textual data and classify messages as fraudulent or non-fraudulent. With a robust preprocessing pipeline, including tokenization and sentiment analysis, the system ensures high accuracy and reliability. By fine-tuning BERT and optimizing model parameters, FraudDetectAI delivers precise predictions and actionable insights.
Project 02:
HomePredictor: Advanced Insights into House Prices
HomePredictor is an interactive web application built using Streamlit, designed for advanced data analysis and prediction of house prices. This project encompasses detailed exploratory data analysis (EDA), feature engineering, and the implementation of XGBoost, a Gradient Boosting Machine Learning model to provide accurate price predictions. Users can explore the relationships between key factors affecting house prices, visualize trends, and predict property values based on input features.
.png)
Project 03: EventShala - Event Management App
Developed a full-stack web application using with Django serving a REST API with React as the frontend. Implemented Payment-Gateways for seamless transactions and Django-JWT for secure user authentication. Included category filters for enhanced event discovery. Users can access detailed event information through their profiles. Robust options for creating, deleting, and tracking events order.
.png)
Project 04: FaceVision
FaceVision is an interactive web application designed for real-time face detection and image processing, built using Python, OpenCV, and Streamlit. The app enables users to upload images, detect faces using Haar cascade classifiers, and apply various filters like grayscale, blur, and edge detection to enhance their images. Users can visualize results instantly through a clean and user-friendly interface, with the added functionality to download the processed images.
.png)