far.ai - Technical Program Manager, Research

• Take complex technical programs from idea to delivery: scope the plan and milestones, then keep complex, multi-team work moving efficiently, transparently, and to a high technical bar. • Prioritize across competing demands in partnership with research teams so the teams’ effort lands on the highest-impact work. • Build and maintain compute and resource roadmaps by surfacing bottlenecks and trade-offs, finding ways to accelerate progress and build relationships with potential external partners. • Partner with researchers to package their work effectively to get our research into the hands of people who can act on it by transferring findings, methods, and tools to frontier labs and other stakeholders. • Connect research and engineering with our internal partners in Operations, Events and Communications giving each the timelines, results, and context they need to plan, resource, and communicate our work. • This role is a great fit if you • Are a fast learner and comfortable with ramping up quickly on unfamiliar technical domains to hold your own in discussions with researchers and engineers. • Can reason about technical trade-offs in depth (e.g. measuring quality of agent output balanced against credit spend, compute efficiency, and training and inference infrastructure) to facilitate clear decisions by stakeholders. • Enjoy 0-to-1 work by building the processes, systems, and structure where none exist, rather than inheriting them. • Thrive in an ambiguous, fast-paced environment that is scaling quickly. • Hold a deep commitment to ensuring safe and beneficial AI. • This role is a poor fit if you • Have no technical background. You don't need to code, but you do need to understand AI/ML well enough to hold your own in discussions with researchers and follow the substance of the work. • Want to inherit a mature TPM function rather than help build one. • Are looking for a product management role with a product roadmap, rather than driving programs, processes, and delivery across research and infrastructure.