Must have 8+ years engineering experience with production systems, including time leading or tech-leading a team. Experience building a team from the ground up - first hires, culture, hiring bar. Shipped AI/LLM products to production - not just demos or POCs. Strong Python and/or TypeScript. Deep hands-on experience with LLMs, agent frameworks (LangGraph, Mastra, AWS Strands, Vercel AI SDK, or similar), prompt engineering, RAG, and model behavior in production. Distributed systems fundamentals: workflow orchestration, fault-tolerant architectures, async patterns. Cloud and self-hosted model deployment (Bedrock, Vertex AI, Azure OpenAI, Anthropic, Ollama). Hands-on leadership - you write code, review PRs, set the bar. You also know when to step back. Comfortable with ambiguity and the pace of a zero-to-one build.