Generative AI & Agents: LLMs/multimodal models, agentic patterns (function calling, reasoning loops, autonomous error correction), frameworks (LangGraph/LangChain/LlamaIndex), vLLM/PyTorch/HuggingFace, production multi-agent systems, inference optimization, MLOps, model monitoring/reliability, ethical AI/compliance. Define architecture for planning/optimization (SAT/SMT, constraint programming, PDDL/planners) and generative AI / agentic systems (LLMs, multimodal models, multi-agent orchestration, inference optimization).