Member of Technical Staff, Pretraining

About the Company

Mirendil is a tech-first company focused on solving core bottlenecks that unlock step-change acceleration across science and technology. We are building a frontier AI research company and training our own models end-to-end, with a team including researchers and engineers from Anthropic, Google DeepMind, xAI, OpenAI, Microsoft, Apple, and MIT.

Responsibilities

• Implement and iterate on model architectures, training algorithms, and optimizer research in large-scale pretraining runs
• Scale distributed training jobs across thousands of GPUs
• Optimize training throughput for novel attention mechanisms, architecture variants, and compute efficiency improvements
• Design and build large-scale data pipelines for efficient model consumption and dataset curation
• Run and analyze scientific experiments to advance understanding of how architecture and data choices affect model capabilities

Requirements

• Experience working at the intersection of research and systems
• Ability to contribute across the full pipeline from data processing to distributed training infrastructure
• Interest in large-scale model pretraining and low-level optimization