Large Language Model Operations (LLMOps)

Running LLMs in production, not just training them

Duke University's LLMOps Specialization is a beginner-level program that fills a gap most generative AI courses skip: what it actually takes to deploy, manage and optimize large language models once they're built, not just how to train one.

What it covers

• Working directly with the OpenAI API and Hugging Face tooling
• Retrieval-augmented generation and fine-tuning in practice
• Prompt engineering and prompt patterns as operational tools
• Responsible AI practices for models running in real systems

It suits developers or technical teams responsible for keeping LLM-powered products running reliably, rather than data scientists focused purely on research. It's a smaller, specialist program with around 18,500 learners so far — reflecting how new the LLMOps discipline itself still is.