Developed a new 1D upsampling operation for the FlagOpen/FlagGems repository, enhancing support for 1D tensor data pipelines. The work focused on implementing the upsample_nearest1d operator using Python and PyTorch, with careful attention to performance benchmarking and tensor manipulation. Addressed numerical precision challenges and ensured robust handling of tensor shapes, resulting in more reliable and accurate processing of 1D data. Comprehensive unit tests were added to validate the new functionality and cover edge cases, supporting maintainability and correctness. This contribution improved the library’s capabilities for 1D data processing and demonstrated a methodical approach to operator development and testing.