PhD student at CNRS

My current research focuses on in-memory representations of heterogeneous and hierarchical data structures for HPC / HTC.

I am particularly interested in parallel computing, low-level optimization, compile-time programming, and building fast, robust libraries. I try to actively participate in C++ library development aimed at providing zero-overhead abstractions for scientific codes.

• Kumi - A C++20 library implementing product types and algebraic operations on them. Built to minimize compiler overhead while providing a pre-C++26 reflection-like interface for record types, acting as a lightweight, faster alternative to std::tuple.
• Kiwaku - A C++23 library designed to handle multidimensional, hierarchical, and heterogeneous data structures with a focus on hardware cache layout and performance.
• EVE - A C++20 SIMD library for portable vectorization across x86, ARM, RISC-V, and PowerPC. Contributed to MSVC compiler compatibility, and currently trying to work on reducing compile-time overhead, and lower memory footprint.

• C++ (Modern C++, since C++20)
• C (Low-Level Programming & High Performance Codes)
• Fortran (beginer)
• Shell / Bash (Scripting & Automation)
• Python (Data Analysis & Tooling)

• Ph.D. in Computer Science - In-memory representations of hierarchical data structures for HPC/HTC
• Master's Degree in High Performance Computing & Simulation (HPCS)
• Bachelor's Degree in Computer Science

• Low-Level Programming & SIMD Vectorization
• Micro-architectural & Memory Optimization
• Compile-Time programming
• Numerical Simulation & Heterogeneous Computing
• Type Theory & Category Theory