My background spans the full AI stack: I've shipped computer vision systems across 600+ cameras at Avathon, designed 27-agent orchestration frameworks for factory-floor AI, and engineered LLM + RAG layers that replaced manual configuration cycles. I've worked across computer vision, multimodal models (LLaVA, Qwen VLMs), multi-agent systems (CrewAI), and edge inference with hardware acceleration (Intel VAAPI, CUDA, DeepStream).