Member of Technical Staff, Infrastructure and Training Systems

About the role

Radical Numerics is seeking a Member of Technical Staff to design and build large‑scale model training infrastructure for biological world models. The role bridges deep systems engineering with modern machine learning to accelerate scientific discovery.

Key responsibilities

• Design, scale, and optimize distributed training systems for large biological models.
• Develop performance optimizations across the stack, including communication patterns, memory efficiency, custom kernels, and instrumentation.
• Create reusable training frameworks and internal libraries to improve reproducibility and reliability.
• Establish standards for robustness, fault tolerance, checkpointing, monitoring, and incident analysis.
• Collaborate closely with researchers, training scientists, and infrastructure engineers to unblock experiments and support new scientific directions.

Required profile

• Strong engineering track record in distributed systems, high‑performance ML infrastructure, or related areas.
• Proficiency in building performant, maintainable software using Python, PyTorch, Triton, CUDA, and C++.
• Deep understanding of modern deep‑learning frameworks and their system internals.
• Ability to debug complex, multi‑layered distributed training systems and performance regressions.
• Comfort working collaboratively with researchers and engineers, with a bias toward initiative and execution.
• Excellent written and verbal communication bridging technical and scientific domains.

Required skills

• Python
• PyTorch
• Triton
• CUDA
• C++

What we offer

• Competitive compensation and comprehensive benefits.
• Support for continual learning and professional development.
• Opportunity to work on frontier systems problems with real biological applications.
• Collaborative culture valuing rigor, creativity, and cross‑disciplinary partnership.