Accenture is hiring an LLM Operations Engineer in Kolkata to build multi-agent AI systems. You will develop applications on large language models such as GPT, Claude and Llama using prompt engineering and orchestration, implement agentic workflows (planner, executor, critic and memory loops), build RAG pipelines on vector databases, and create tool-using agents that connect with APIs, databases and enterprise systems. The role also covers deploying AI copilots, tuning LLM performance with fine-tuning, prompt chaining and caching, and applying MLOps/LLMOps practices for monitoring, evaluation and guardrails, with a focus on responsible AI. Useful experience includes agent planning methods (ReAct, Plan-and-Execute), LLM evaluation tools (RAGAS, TruLens, Promptfoo) and knowledge graphs. The posting asks for 3-8 years of experience.