Wafer AI - Member of Technical Staff
Core
Autonomous AI performance engineers optimizing inference on heterogeneous hardware to maximize intelligence per watt.
Role type
Member of Technical Staff (Systems/Inference Optimization)
Builds
Production inference serving stacks on multi-vendor clusters
Domain
AI Infrastructure / Heterogeneous Hardware
Deliverable
production ML models
Required skills
CUDA kernel development, HIP kernel development, Triton kernel development, inference serving optimization, batching strategies, KV cache optimization, speculative decoding, quantization, cluster design and operation, observability, cost optimization
Preferred skills
Experience with NVIDIA, AMD, TPU, AWS Trainium, D-Matrix
Technologies
CUDA, HIP, Triton, Python, TPU, AWS Trainium
Responsibilities
Ship day-zero support for new open-source models tuned for latency/throughput, Optimize serving stack including batching, KV cache, speculative decoding, and quantization, Write and tune kernels in CUDA, HIP, and Triton for various hardware, Design and operate heterogeneous multi-vendor clusters, Run production inference focusing on reliability, observability, and cost per token
Seniority
Mid-level (1-6 years experience)