Pedro Maries
Principal Engineer - AI And Modelling at Allfunds
Turin, Piedmont, Italy
Summary
👤
Senior🎓
Top SchoolPedro Maries is a Principal Engineer specializing in numerical software and AI-driven modelling, bringing a decade of hands-on experience building end-to-end systems from architecture and high-performance algorithm design to deployable artifacts. Based in Turin, he favors pragmatic engineering—Test-Driven Development, containerization, and using the language that gets the job done—often Python for agility, C for performance insight, and increasingly Go for concurrent services. At Allfunds he leads AI and modelling efforts, applying aerospace-trained analytical rigor (BSc+MSc, top 7% from Universidad Politécnica de Madrid) to practical engineering problems. An active open-source contributor, he improved critical numerical functionality in the widely used Numba project by implementing and hardening np.cross with broadcasting and comprehensive tests. He’s pragmatic, curious, and prefers Twitter or email over LinkedIn for technical conversations, reflecting a builder’s focus on craft over profile.
10 years of coding experience
BSc + MSc, Aerospace Engineering, Top 7%, BSc + MSc, Aerospace Engineering, Top 7% at Universidad Politécnica de Madrid
Italian, Spanish, English
Stackoverflow
91reputation
8kreached
2answers
0questions
Github Skills (10)
compiler10
compiler-compiler10
numba10
python10
numpy10
testing9
llvm7
google-cloud-firestore6
cookiecutter6
parallel5
Programming languages (7)
Github contributions (5)
numba/numba
May 2019 - Oct 2019
NumPy aware dynamic Python compiler using LLVM
Role in this project:
Backend Developer
Backend DeveloperContributions:38 commits, 2 PRs, 28 comments in 4 months
Contributions summary:Pedro focused on implementing and testing the `np.cross` function within the Numba compiler, a core feature for numerical computation. Their contributions included the initial implementation, breaking it down into different cases based on input array dimensions, and incorporating broadcasting to handle higher-dimensional arrays. The user added several test cases to ensure the function's correctness and exception handling. They also refactored the code by extracting common operations, which improved code organization and readability.
cudapythonparallelnumpynumba
partmor/SDC-Traffic-Sign-Classifier
Apr 2017 - Jul 2017
Contributions:27 commits, 25 pushes, 1 branch in 3 months
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.