senior-ml-engineer

To set up an ML deployment pipeline, you need to export, containerize, and deploy models to staging and production. This skill provides dedicated scripts to bootstrap this workflow and enforce production-grade requirements like automated testing.

Monitoring ML drift requires MLOps instrumentation that tracks data and model performance changes over time. This skill provides reference implementations for drift monitoring, A/B tests, and automated retraining to maintain model reliability.

Building a RAG system involves creating pipelines that retrieve relevant context for large language models. You can use the RagSystemBuilder script provided here to bootstrap production-ready RAG workflows and integrate scalable LLM patterns.

Containerized environments are required for staging canaries and automated testing to ensure reproducible infrastructure. This skill enforces production-grade requirements by providing reference templates for reliable, scalable model deployments.

Automated retraining pipelines can be triggered when drift monitoring detects significant data or model degradation. This skill includes MLOps patterns to orchestrate automated tests, evaluate canaries, and execute retraining workflows.

Production-grade ML patterns are needed when operationalizing models and LLMs that require high reliability, drift detection, and scalable infrastructure. Use these patterns to move from experimental scripts to robust, monitored production environments.