Tokens-as-a-Service (Taas) Software Engineer

📋 Description

• • As a Tokens-as-a-Service (TaaS) Software Engineer at OpenAI, you will play a critical role in transforming raw compute capacity into measurable, reliable token throughput for AI workloads, directly enabling the performance and scalability of OpenAI’s models across first-party and partner infrastructure.
• • Your day-to-day responsibilities will include developing systems and tooling to measure, monitor, and improve token throughput; supporting performance benchmarking and tokenomics analysis; building integrations for external compute environments; monitoring operational metrics like utilization, SLAs, and billing; identifying bottlenecks in hardware, networking, and software; and collaborating with cross-functional teams to translate infrastructure capacity into productive AI workloads.
• • You will join a highly technical, mission-driven team at OpenAI focused on pushing the boundaries of AI infrastructure, working alongside experts in compute, networking, finance, and operations to ensure that GPU capacity is efficiently onboarded, measured, and optimized for real-world model training and inference.
• • In this role, you will deepen your expertise in distributed systems, performance engineering, infrastructure observability, and AI workload optimization, while contributing to systems that directly impact the efficiency, reliability, and scalability of one of the world’s leading AI research organizations.