Over a two-month period, JPS Amaroo developed GPU-accelerated scientific simulations and improved documentation across JuliaParallel/julia-hpc-tutorial-sc24 and JuliaLang/www.julialang.org. He implemented configurable stencil and reaction-diffusion models using Julia, CUDA, and KernelAbstractions, enabling multi-backend support and performance demonstrations for parallel computing workflows. His work included setting up Buildkite-based CI pipelines for CUDA testing across Julia versions, ensuring code stability and maintainability. Additionally, he enhanced educational content by refining GPU programming tutorials and cleaned up outdated high-performance computing documentation, aligning it with upcoming Dagger scheduling features. The work demonstrated depth in scientific computing, CI/CD, and documentation management.