Member of Technical Staff, Inference

About the Company

Radical Numerics is an AI research lab building general biological intelligence. Our mission is to master the code of life and reduce human suffering through generative genomics and frontier biological AI models.

Responsibilities

• Drive end-to-end performance improvements across the inference stack, including model execution, memory management, and hardware utilization
• Develop high-performance inference primitives, GPU kernels, and numerical operators
• Partner with pharmaceutical companies, biotech organizations, and government partners to deploy models in production
• Build scalable deployment infrastructure for serving, monitoring, and benchmarking foundation models
• Collaborate with research and platform teams to translate AI research into real-world impact

Requirements

• Proven experience optimizing and operating LLMs or foundation models in production
• Deep understanding of GPU architectures and experience with CUDA or Triton
• Strong software engineering fundamentals with proficiency in Python and PyTorch
• Ability to diagnose bottlenecks across model architectures, networking, and distributed infrastructure
• Experience working with customers or cross-functional stakeholders to deliver production systems

Preferred Qualifications

• Experience with inference frameworks like vLLM, TensorRT-LLM, SGLang, or DeepSpeed
• Contributions to open-source AI infrastructure, compilers, or kernel libraries
• Experience with distributed systems and large-scale GPU clusters
• Familiarity with computational biology, genomics, or scientific AI
• Experience operating AI systems in regulated or mission-critical environments