Senior/Principal AI Performance Engineer

📋 Description

• Design and fine-tune inference pipelines for LLMs to maximize throughput and minimize latency.
• Apply quantization, pruning, speculative decoding, batching, and kernel fusion.
• Optimize inference-serving stacks (vLLM, TensorRT-LLM, ONNX Runtime) for production.
• Profile and tune GPU/accelerator utilization across the full inference stack.
• Design and optimize distributed training pipelines for large-scale models.
• Tune training efficiency with mixed-precision, checkpointing, and optimizer improvements.

🎯 Requirements

• Deep expertise in LLM inference optimization, including serving frameworks (vLLM, TensorRT-LLM, ONNX Runtime).
• Strong background in distributed AI training with PyTorch FSDP, DeepSpeed, Megatron-LM, or JAX/XLA.
• Experience packaging AI environments for reproducible deployment (containers, Apptainer/Singularity).
• Fluency with GPU profiling tools: NVIDIA Nsight, PyTorch Profiler, CUDA analysis.
• Familiarity with HPC environments: Slurm, PBS, RDMA/InfiniBand, MPI.
• Experience integrating AI workloads into CI/CD pipelines with automated testing and benchmarking.
• Comfort using LLM-based tools and agentic frameworks; strong analytical and communication skills.

🎁 Benefits

• Medical, dental, and vision insurance.
• Flexible paid time off.
• Employee stock options.
• Remote work; no travel required for most positions.