Solid hands-on experience designing, shipping, and operating LLM-powered agents in production, with ownership of output correctness, grounding, and evaluationExperience building Graph RAG retrieval and agent memoryAbility to assemble context that respects each user’s entitlements and access controlsExperience building evaluation harnesses and guardrails for LLM agentsExperience composing multi-agent workflows using A2A (agent-to-agent) and MCP (Model Context Protocol)Experience owning the models behind the agents, such as next-best-action and retrieval ranking for a copilot, and price optimization and elasticity for pricingUnderstanding of how to design agents that explain and coordinate while relying on deterministic services for prices and approvals, so agents never invent a numberStrong communication skills to work with product, platform, and pricing stakeholders