Build and extend a production LLM agent platform: design agents and tools, implement integrations, own RAG ingestion and retrieval, define integration contracts, drive evaluation-led development, operate deployments and observability, and establish AI-specific engineering practices across frontend and backend.