Software Engineer, Model Inference

Highlights

• Bring the world's largest AI models to production at high volume and low latency.
• Improve inference performance, throughput, and efficiency across the model stack.
• Partner with researchers to accelerate progress via fast, reliable inference.

What they offer

• Equity

Ideal candidate

Self-directed engineer who owns problems end-to-end and obsesses over performance and reliability. Humble, collaborative builder excited to help researchers ship frontier AI models to the world.

Requirements

• At least 5 years of professional software engineering experience
• Understanding of modern ML architectures and intuition for optimizing inference performance
• Familiarity with PyTorch, NVIDIA GPUs, NCCL, CUDA, InfiniBand, MPI, and NVLink
• Experience architecting, building, observing, and debugging production distributed systems, ideally performance-critical
• Experience rebuilding or refactoring production systems for rapidly increasing scale
• End-to-end ownership, self-directed, humble and collaborative