Mani khandan

ML Engineer

• Job Details:Developed and deployed machine learning models to production, focusing on predictive analytics, classification, and real-time inference use cases. Preprocessed large-scale datasets with efficient feature engineering, integrating feature store concepts to ensure consistency across training and serving. Designed and automated scalable data pipelines for training and evaluation using Python, Scikit-learn, and TensorFlow, improving workflow reproducibility. Built end-to-end ML workflows with data preprocessing, training, and model registry integration for production-ready deployment and model tracking. Conducted A/B testing and model evaluation using metrics like accuracy, precision, recall, F1-score, and ROC-AUC to ensure high model performance in production. Tuned hyperparameters and applied continuous training strategies to improve accuracy, reduce model drift, and maintain robust inference over time. Leveraged MLflow for experiment tracking, model versioning, and registry, ensuring reproducibility and traceability across the ML lifecycle. Implemented monitoring and logging using Prometheus and Grafana, enabling live performance tracking, anomaly detection, and metric-based model promotion. Integrated CI/CD pipelines using GitHub Actions and Docker for automated model testing, containerization, and deployment across cloud environments. Collaborated cross-functionally with data engineers and product teams to align ML solutions with business objectives, reducing model deployment time by 30%.