About this role
The Sr. SW Test Engineer - Neural Library will be responsible for validating neural network operators and libraries for a next-generation AI computational storage product. The role focuses on ensuring the correctness, performance, and robustness of compute kernels deployed on advanced hardware platforms. Responsibilities include validating operators like matmul, convolution, and attention against reference models, developing automated tests for operator and graph-level validation, and performing numerical accuracy checks against frameworks such as PyTorch or TensorFlow. The engineer will also execute smoke and regression test suites, benchmark performance metrics including throughput and latency, and conduct software profiling to identify execution bottlenecks and memory leaks. Collaboration with compiler, runtime, and hardware teams is essential to debug failures across the execution stack. Required qualifications include 5 years of experience (based on the 5-8 year range provided) in AI/ML systems validation or performance testing, strong C++ programming skills, proficiency in Python for test automation, and a deep understanding of deep learning operators and frameworks. Preferred qualifications include experience with profiling tools like perf, VTune, or Nsight, familiarity with ML execution stacks such as MLIR/XLA, exposure to accelerator validation (GPU/NPU/FPGA), and experience validating large language model workloads.