Job Description

xAI is seeking a highly motivated engineer to optimize model inference, build production serving systems, and accelerate research on scaling test-time compute. The role involves hands-on contributions to open-source projects and requires strong communication and prioritization skills.

Responsibilities

• Optimizing latency and throughput of model inference
• Building reliable and performant production serving systems
• Accelerating research on scaling test-time compute
• Model-hardware co-design for next-generation architectures
• System optimizations for model serving (batching, caching, load balancing)
• Low-level optimizations for inference (GPU kernels, code generation)
• Algorithmic optimizations for inference (quantization, distillation)
• Working on large-scale inference engines

Requirements

• Experience with Python/Rust, PyTorch/JAX, CUDA/CUTLASS/Triton/NCCL, Kubernetes
• Experience with system optimizations for model serving
• Experience with low-level optimizations for inference
• Experience with algorithmic optimizations for inference
• Experience with large-scale inference engines or reinforcement learning frameworks
• Experience with large-scale, high-concurrent production serving
• Experience with testing, benchmarking, and reliability of inference services
• Strong communication skills

Benefits

• No benefits