What problem does it solve? Moving machine learning models from notebooks to production requires expertise in deployment pipelines, monitoring, scaling, and LLM integration that most teams lack in one place. ## Core Features & Use Cases - Model Deployment Pipelines: Structure model serving with low-latency inference, A/B testing, feature stores, and automated retraining using Docker, Kubernetes, and cloud platforms. - RAG & LLM Integration: Build retrieval-augmented generation systems and integrate LLMs with LangChain, LlamaIndex, and vector databases like Pinecone. - ML Monitoring & Observability: Set up drift detection, health checks, and performance tracking with MLflow, Weights & Biases, and Prometheus. - Use Case: A team needs to deploy a PyTorch model behind an API with 99.9% uptime, monitor for data drift, and add a RAG-based chatbot—this Skill provides the architecture patterns, reference guides, and scaffolding scripts for each stage. ## Quick Start Ask the AI to design a production deployment pipeline for a PyTorch model with monitoring and drift detection using this skill.