Multi-agent Orchestration at Depth: Hands-on experience designing, running, and scaling multi-agent systems (e.g., Claude Code, Codex, Cursor background agents, custom orchestrators, LangGraph-style graphs) - including MCP tooling, shared context and memory, agent handoffs, and robust eval harnesses. Cross-functional Leadership: Partner with Product, Security Research, UX, and Support leaders to translate ambiguous customer needs into concrete, shippable technical plans; conduct deep design reviews; and raise the quality bar through thoughtful mentorship.