Own the infrastructure powering Pax Historia’s AI requests across 37+ models and multiple providers. Responsibilities include improving provider reliability, structured outputs, caching, routing, monitoring, cost efficiency, latency, and scalability for billions of monthly tokens. This founding engineer role requires strong fundamentals, attention to detail, data-driven decision-making, and rapid learning. The position is fully in-person in San Francisco and does not involve model training or self-hosted inference.