Production-grade machine learning pipeline for customer churn prediction with modular components for data preprocessing, model training, and streaming inference. This third repo in a series builds on previous work to create a deployment-ready prediction system with comprehensive configuration, evaluation metrics, and scalable architecture.

Computer vision system that identifies dog breeds with 93.3% accuracy using transfer learning (VGG, ResNet, AlexNet). Features model comparison, performance analytics, and applications for veterinary tech, pet retail, and animal shelters.