About the service
I am a Fortran developer specializing in scientific and high-performance computing, and I have spent many years inside numerical code that others are afraid to touch. My daily work is writing fast and correct simulations, solvers, and data processing pipelines in modern Fortran (from F77 to 2008/2018) and making them run efficiently on multi-core machines and clusters. Unlike those who reach for Julia, MATLAB, or C++ by habit, I consciously stay in the Fortran world: for dense numerical loops, linear algebra, and legacy scientific software, it remains the most honest and predictable tool. I care about numerical stability, reproducibility, and correct physics, not just that the code compiles.
Much of my work is rescue and modernization. Research groups and engineering teams often have Fortran code decades old: it still produces results everyone relies on, but no one remembers how it works. I carefully refactor such code, add tests and documentation, replace fixed-format spaghetti with clean modules, integrate it into build systems and CI without changing the science. When speed is the goal, I parallelize through OpenMP and MPI, tune memory access and vectorization, and honestly profile to see where time really goes. When needed, I link Fortran cores with Python or C for tools and visualization.
If you have a solver that's too slow, a simulation that needs to scale, or an old scientific program that has years ahead of it, I can help. I explain tradeoffs in plain language, keep results verifiable at every step, and hand over code your team can maintain without me.
— Development in modern Fortran (F77 to 2008/2018), modules and clean interfaces
— HPC and parallel computing via OpenMP and MPI on clusters
— Numerical methods: linear algebra, ODE/PDE solvers, integration, optimization
— Simulations and modeling for physics, engineering, and applied mathematics
— Legacy code rescue: refactoring, testing, build, documentation
— Performance profiling, vectorization, and memory tuning
— Linking Fortran with Python and C for tools and visualization
Much of my work is rescue and modernization. Research groups and engineering teams often have Fortran code decades old: it still produces results everyone relies on, but no one remembers how it works. I carefully refactor such code, add tests and documentation, replace fixed-format spaghetti with clean modules, integrate it into build systems and CI without changing the science. When speed is the goal, I parallelize through OpenMP and MPI, tune memory access and vectorization, and honestly profile to see where time really goes. When needed, I link Fortran cores with Python or C for tools and visualization.
If you have a solver that's too slow, a simulation that needs to scale, or an old scientific program that has years ahead of it, I can help. I explain tradeoffs in plain language, keep results verifiable at every step, and hand over code your team can maintain without me.
— Development in modern Fortran (F77 to 2008/2018), modules and clean interfaces
— HPC and parallel computing via OpenMP and MPI on clusters
— Numerical methods: linear algebra, ODE/PDE solvers, integration, optimization
— Simulations and modeling for physics, engineering, and applied mathematics
— Legacy code rescue: refactoring, testing, build, documentation
— Performance profiling, vectorization, and memory tuning
— Linking Fortran with Python and C for tools and visualization
