In November 2025, Hahn Se developed a new interface for JuliaGPU/CUDA.jl that enables batched eigenvalue computations on 3D StridedCuArray inputs using Julia. The work focused on extending the XsyevBatched! interface to support 3D matrix workflows, addressing the need for efficient large-scale linear algebra operations on GPUs. Hahn implemented cuSOLVER version compatibility checks to ensure reliable execution across different GPU environments and introduced memory-aware buffer size calculations to optimize resource usage. This feature leveraged GPU programming and numerical computing expertise, providing a robust solution for batched eigenvalue problems and demonstrating depth in both performance optimization and API design.