Build and own production-grade agentic AI systems end to end, including multi-agent orchestration, LLM and vision-language model integration, retrieval, memory, guardrails, evaluation, observability, and deployment pipelines. The role requires designing reliable systems for cloud, on-premises, air-gapped, and offline environments; handling large multimodal datasets; and setting technical direction for other engineers. Strong Python, software engineering, production AI/ML, and secure deployment experience are required.