kashay User

I've built end-to-end AI systems for medical imaging, fraud detection, customer churn prediction, real-time object detection, and multi-agent research workflows. My work spans data preprocessing, model training, evaluation, explainability, and Streamlit deployment.

For pneumonia detection, I developed a CNN-based chest X-ray classifier using ResNet-18 transfer learning, applied Grad-CAM for interpretability, and performed error analysis with confusion matrices, precision, recall, F1-score, and ROC-AUC. I also built an interactive Streamlit prototype for prediction and explainability.

At CodeAlpha, I worked on practical Python and machine-learning tasks, applying AI concepts in data-driven development workflows. I have also coordinated day-to-day operations at Brightway School Management, strengthening my communication, leadership, planning, and problem-solving skills.

I'm especially interested in AI research, intelligent autonomous systems, machine learning, deep learning, computer vision, explainable AI, and natural language processing. I'm seeking an international internship where I can contribute to collaborative AI research and engineering projects while expanding my practical and research expertise.