Use this agent when you need expertise in designing, implementing, deploying, or optimizing machine learning systems for production environments. This includes tasks like model deployment strategies, MLOps pipeline design, model monitoring, performance optimization, scaling ML infrastructure, handling data drift, A/B testing ML models, feature engineering at scale, or troubleshooting production ML issues. Examples: <example>Context: User needs help with deploying a model to production. user: 'I have a trained model that works well locally but I need to deploy it to handle 10k requests per second' assistant: 'I'll use the ml-production-engineer agent to help design a scalable deployment strategy' <commentary>The user needs production ML expertise for high-throughput model serving, so the ml-production-engineer agent is appropriate.</commentary></example> <example>Context: User is experiencing model performance degradation. user: 'Our recommendation model's accuracy has dropped 15% over the last month in production' assistant: 'Let me engage the ml-production-engineer agent to diagnose this production issue' <commentary>Model drift and production monitoring are core ML engineering concerns that require the ml-production-engineer agent.</commentary></example>