An end-to-end AI-powered surveillance system for public safety monitoring with multi-source data fusion, temporal validation, and vision-language scene interpretation.

AI-powered crop monitoring platform that helps farmers detect plant diseases, pests, and nutrient deficiencies before they cause major losses. Using drone imagery and deep learning, our system scans fields in real time and delivers instant, actionable insights through a simple web dashboard 

Applied advanced data analytics including geospatial mapping and machine learning to analyze transportation incidents. Used a Decision Tree model to identify key risk factors and deliver actionable safety and logistics recommendations.

Developed machine learning models to predict project milestone success using five datasets. Applied data preprocessing, statistical evaluation, hyperparameter tuning, and LIME for model interpretability.

Built linear and logistic regression models to predict housing prices and classify personal loan acceptance, evaluating performance using R², MSE, RMSE, accuracy, F1 score, and confusion matrix.

Built a hotel management system with a Java Swing UI and relational database, managing hotels, rooms, guests, reservations, and services. SQL queries handled room availability, active reservations, and guest billing.

Implemented and compared SVM, XGBoost, Random Forest, and CNN on Fashion MNIST, evaluating accuracy, F1 score, and training efficiency. CNN achieved the highest accuracy while XGBoost excelled in efficiency, highlighting trade-offs between complexity and performance.

Analyzed multiple datasets using Python and Pandas, applying outlier detection, statistical analysis, and data visualization techniques including histograms, scatter plots, and correlation matrices. 

Built an event-driven network simulator in Java modeling routing, congestion control, and security protocols. Implemented routers, switches, and Dijkstra’s algorithm for shortest-path routing to evaluate network performance.