Built a robust predictive system to identify at-risk customers. Utilized ensemble methods including Random Forest and XGBoost on a dataset of over 1 million user records to forecast churn probability with high accuracy.

Developed a Deep Learning model using Convolutional Neural Networks (CNNs) to detect anomalies in X-ray images. Achieved 94% accuracy and deployed the model as a RESTful API service for real-time inference.

Won first place in college hackathon for developing a rapid, high-impact technical solution under strict time constraints.