Developed machine learning models for anomaly detection and predictive maintenance, reducing accident risks by identifying track faults, equipment wear, and potential issues, resulting in improved railway safety and proactive maintenance scheduling.Utilized Python, TensorFlow, and scikit-learn to build and fine-tune classification and clustering models, including random forests, SVM, and k-means, improving fault detection accuracy by 25%.Implemented Apache Spark and Hadoop for distributed processing of large datasets, and integrated MongoDB for efficient storage and retrieval of unstructured data across multiple subsystems.Built time-series forecasting models for train arrival prediction and passenger flow estimation using ARIMA and LSTM networks, optimizing railway scheduling and resource management, which reduced delays by 15%.Applied computer vision techniques using OpenCV and CNNs for real-time object detection and video analysis, enhancing safety by detecting obstacles, intrusions, and equipment damage with a 90% accuracy rate.Deployed models using Docker and orchestrated with Kubernetes for scalable cloud-based operations, leveraging AWS services such as S3, EC2, and Lambda to ensure efficient and fault-tolerant model deployment.Created automated CI/CD pipelines with Jenkins for continuous integration and deployment, and utilized Terraform and Ansible to automate infrastructure management, improving deployment efficiency by 40%.Collaborated closely with cross-functional teams, ensuring seamless integration of machine learning models into real-time railway monitoring systems, contributing to overall operational improvements.