Member of Technical Staff - Pre-Training Infra

Reflection is a research lab building open AI models. The Member of Technical Staff will build and scale distributed infrastructure for foundation-model pre-training, optimize large training workloads, and collaborate with researchers to turn experimental ideas into production-ready systems.

Responsibilities:

• Build and scale distributed training systems that power frontier model pre-training
• Work closely with research teams to design and operate large-scale training runs for foundation models
• Develop infrastructure that enables efficient training across thousands of GPUs using modern distributed training frameworks
• Optimize training throughput, stability, and efficiency for large model training workloads
• Collaborate directly with pre-training researchers to translate experimental ideas into scalable, production-ready training systems
• Improve performance of distributed training workloads through optimization of communication, memory usage, and GPU utilization
• Build and maintain training pipelines that support large-scale datasets, checkpointing, and experiment iteration
• Debug and resolve performance bottlenecks across distributed training stacks including model parallelism, GPU communication, and training runtime systems
• Contribute to the development of systems that enable rapid experimentation and iteration on new training techniques