About the job
Our Mission
At Reflection AI, our mission is to develop open superintelligence and ensure its accessibility to everyone.
We are pioneering open weight models for individuals, agents, enterprises, and even nations. Our dynamic team comprises AI researchers and innovators from top-tier organizations like DeepMind, OpenAI, Google Brain, Meta, Character. AI, Anthropic, and others.
About the Role
As a Research Program Manager at Reflection, you will serve as a crucial leader and facilitator, working closely with our research and engineering teams to expedite the development of advanced models. Rather than merely tracking projects, you will act as a force multiplier, clarifying uncertainties, driving decision-making, and ensuring cohesive collaboration across teams.
This position integrates directly with our Pre-training Machine Learning and Data teams, collaborating with our research leads to shape the future of frontier models. You will engage deeply with the research lifecycle, from coordinating data pipelines and planning experiments to guiding model architecture decisions and implementing scaling strategies. Your role will involve identifying key leverage points for improvement and establishing the processes that allow researchers to focus on innovation rather than organizational challenges.
You will embody a proactive mindset. In challenging situations, you will not wait for directives; instead, you will assess the circumstances, streamline communication, ensure alignment, and drive towards resolutions.
What You'll Do
Integrate into the Pre-training ML and Data teams to gain a comprehensive understanding of the technical environment, foster trust with researchers and technical leads, and pinpoint where processes can significantly enhance research efficiency.
Oversee the implementation of complex, cross-functional research initiatives that involve data management, model architecture, training sessions, and evaluations, often in the absence of predefined guidelines.
Manage the operational flow of pre-training research, including prioritizing experiments, scheduling runs, ensuring data readiness, and facilitating handoffs to downstream teams.
Empower research leadership to make swift decisions by deeply analyzing technical trade-offs and presenting clear, actionable recommendations.
Design streamlined processes that introduce structure to the fluid nature of research environments without adding unnecessary barriers.
