At HEP experiments, processing billions of records of structured numerical data can be a bottleneckin the analysis pipeline. This step is typically more complex than current query languages allow,such that numerical codes are used. As highly parallel computing architectures are increasinglyimportant in the computing ecosystem, it may be useful to consider how accelerators such as GPUscan be used for data analysis. Using CMS and ATLAS Open Data, we implement a benchmarkphysics analysis with GPU acceleration directly in Python based on efficient computational kernelsusing Numba/LLVM, resulting in an order of magnitude throughput increase over a pure CPU-based approach. We discuss the implementation and performance benchmarks of the physicskernels on CPU and GPU targets. We demonstrate how these kernels are combined to a modernML-intensive workflow to enable efficient data analysis on high-performance servers and remarkon possible operational considerations.