Network performance engineer optimizing NIC, PCIe, driver, kernel, and RDMA communication for single- and multi-host AI inference clusters — improving KV cache transfer, cross-node collective communication, and model-weight loading, while troubleshooting packet loss and latency. Core stack: Linux, RDMA (RoCEv2/InfiniBand), C/C++, plus Python/Go/Shell.
Sign in to see your match