I build production-grade LLM and agentic AI systems from scratch — RAG pipelines, structured prompt architectures, multimodal AI, FastAPI backends, and full deployment with Docker. Every system I deliver runs in production, not just in notebooks. Scope this includes: