During a two-month period, Pere contributed to ggml-org/llama.cpp, Mintplex-Labs/whisper.cpp, and FreeCAD/FreeCAD, focusing on API design, build system configuration, and post-processing reliability. Pere exposed ggml-cpp.h as a public header in both llama.cpp and whisper.cpp using CMake, enabling external integration and improving modularity for downstream projects. In FreeCAD, Pere addressed a critical issue in the Fanuc post-processor by removing a trailing percent sign from G-code output, preventing controller hangs during drip feeding. The work demonstrated strong proficiency in C++, CMake, and Python, with careful attention to error handling, unit testing, and robust automation workflows.