LLMOps
Definition pending
We've cataloged "LLMOps" but haven't written a full definition yet. Definitions are hand-curated rather than auto-generated, so it takes time to cover every term.
Want this one prioritized? Email us and we'll bump it.
Practical example
LLMOps is MLOps specialized for LLMs — it adds concerns unique to large language models: prompt management, guardrails, context window management, token cost tracking, hallucination monitoring, and RAG pipeline health. LLMs aren't just bigger models; they require new operational practices.
Workflow example
LLMOps deployment: (1) prompt versioning: store prompts in git with the code, (2) guardrail monitoring: track input/output guardrail trigger rates, (3) RAG monitoring: retrieval latency, recall@k, context relevance, (4) cost tracking: monitor tokens per request, cost per request, cost per user, (5) hallucination detection: sample outputs, check factual accuracy, alert if hallucination rate increases, (6) model router: route queries to appropriate model based on complexity — small model for simple, large for complex.