Mingming Xu developed Tile operator support within the microsoft/onnxruntime repository, focusing on the WebNN Execution Provider to expand hardware-accelerated tensor operations. Using C++ and leveraging ONNX Runtime’s extensibility, Mingming integrated the Tile operator path to improve execution efficiency for models with heavy tiling requirements. The implementation reduced reliance on CPU fallback by enabling broader use of WebNN, which can increase throughput on supported devices. The work involved validating the integration for correctness and ensuring readiness for further testing. This contribution deepened the WebNN EP’s operation coverage and demonstrated proficiency in C++, tensor operations, and ONNX Runtime internals.