Build production LLM applications with RAG, prompt engineering, agents, evaluation harnesses, guardrails, and cost/latency control. Use when designing or hardening systems that depend on LLM inference, including chat, search, summarization, and agentic workflows.
