Research Program Manager, Research Infra and Training Frameworks

Reflection • San Francisco, CA

1d

About The Position

Reflection is a research lab building open models to make intelligence accessible for everyone to use, customize, and build on. Research Program Managers embed with research and infrastructure teams to accelerate frontier model development. They bring clarity to ambiguous situations, drive decisions, and connect work across multiple teams into a coherent whole.

This role focuses on scaling research infrastructure for frontier-scale training across pre-training, mid-training, and post-training. The manager will help gather and unify requirements and lead a multi-quarter rewrite and continued evolution of the training stack in partnership with Research and GPU Infrastructure engineers. The work spans training frameworks and workloads that pair the trainer with a high-throughput inference engine for reinforcement learning. The role calls for firsthand experience with GPU infrastructure and judgment about which systems and abstractions work at large scale, including 10,000-plus GPUs. It also requires a first-responder approach to incidents: assessing situations, aligning stakeholders, and driving resolution.

Reflection describes itself as a small, fast-moving, talent-dense team where this role will help establish research program management practices and support the development of open foundational models.

Requirements

• At least 7 years of experience in technical program management, research operations, or infrastructure coordination.
• Deep technical knowledge sufficient to engage with engineers on distributed training frameworks, GPU cluster architecture, scheduler behavior, networking, and storage systems.
• Experience in ML/AI or large-scale distributed systems environments is preferred by the posting's use of “ideally.”
• Ability to operate effectively in high-ambiguity, fast-moving environments and create structure and clarity without waiting for permission.
• Track record managing complex, multi-team programs with competing priorities and hard deadlines, including making and clearly communicating tradeoffs.
• Strong stakeholder management skills across technical individual contributors and senior leadership.
• Ability to remain calm and prioritize under pressure, drive resolution, and follow through.
• Motivation to enable researchers and engineers to build capable open-weight AI systems.

Nice To Haves

• Experience in ML/AI or large-scale distributed systems environments.
• Firsthand experience working deeply with GPU infrastructure teams to build training frameworks, including an understanding of which abstractions work at 10,000-plus GPU scale.

Responsibilities

• Own cross-functional programs spanning training infrastructure across pre-training, mid-training, and post-training workstreams.
• Define and own the roadmap for training frameworks at Reflection.
• Partner with infrastructure and research engineering leads to identify bottlenecks, define priorities, and connect infrastructure investments to research velocity.
• Build and maintain visibility into training run health so leadership and teams can make informed decisions.
• Create lightweight, durable processes for cross-team handoffs, configuration management, checkpoint workflows, and other coordination-heavy touchpoints.
• Translate technical complexity into clear status updates and decision frameworks for engineering leadership and executives.
• Gather and unify requirements and help scope and drive the scale-up rewrite of the training stack with Research and GPU Infrastructure engineers.
• In urgent situations, assess the situation, align relevant stakeholders, and drive resolution.

Benefits

• Top-tier compensation, including salary and equity structured to recognize and retain talent globally.
• Stock options.
• Comprehensive medical, dental, vision, and life insurance, with an annual wellness allowance.
• Lunch and dinner provided in the office daily.
• 22 weeks of paid parental leave for birthing and non-birthing parents, including adoptive and surrogate journeys.
• Unlimited paid time off in the U.S. and 30 vacation days in the U.K.
• Visa sponsorship and support for long-term immigration pathways where applicable.
• Regular off-sites, happy hours, and team celebrations.

What This Job Offers

• Job Type: Full-time
• Career Level: Senior
• Education Level: No Education Listed