# Louis Peyrondet

> SWE @ Magnus Medical | Performance Engineer HPC/AI

Location: San Francisco Bay Area, United States
Profile: https://flows.cv/louispeyrondet

HPC & AI Performance Software Engineer passionate about squeezing every bit of performance out of hardware and algorithms. I like building things from scratch and understanding systems at their deepest level.

My background spans low-level and parallel systems programming, high-level ML, GPU and accelerator programming, and distributed systems, which lets me tackle performance problems across the full stack. I approach optimization through a deep understanding of hardware architecture and the interplay between software design and the underlying silicon, whether that means tuning compute kernels for GPUs, extracting every cycle out of modern CPUs, or scaling workloads across nodes.

Always looking to work on hard performance problems at the intersection of HPC, AI, and systems engineering.

## Work Experience
### Software Engineer @ Magnus Medical
Jan 2025 – Present | San Francisco Bay Area
Working on optimization and algorithms to treat depression :
- Overhauled a large-scale fMRI processing pipeline by migrating it from MATLAB to Python, reducing end-to-end runtime by >60%
- Re-engineered core image processing and neuroimaging algorithms, applying high-performance computing principles achieving 150x speedups
- Developed a new U-Net deep learning model for anatomical landmark prediction, improving average accuracy by over 70%
- Reduced runtime and cloud computing costs by ~50% by profiling and optimizing resource allocation within cloud Kubernetes cluster

### Research Intern, HPC Parallel Filesystem optimization @ Inria
Jan 2023 – Jan 2023 | Bordeaux, Nouvelle-Aquitaine, France
Conception et implémentation d'une version décentralisée de I/O-Sets, une méthode d'ordonnancement I/O pour les systèmes de fichiers des supercalculateurs.

### Research Intern, HPC I/O Optimization @ Inria
Jan 2022 – Jan 2022 | Bordeaux, Nouvelle-Aquitaine, France
Développement d'un simulateur I/O afin d'étudier puis d'améliorer les performances I/O d'une biliothèque d'analyse numérique de méthodes des éléments finis pour une utilisation dans un contexte de Calcul Haute Performance.
Publication disponible sur l'IEEE Xplore et présentation en conférence internationale du travail réalisé.

Compilation et exécution du code sur des supercalculateurs tel que PlaFRIM ou Santos Dumont au Brésil avec l'utilisation de SLURM.


## Education
### 5-year Engineering Degree in Computer Science
ENSEIRB-MATMECA - Bordeaux INP

### Licence 2 in Informatique
Université de Bordeaux

### Semester Abroad in Computer Science
Keio University


## Contact & Social
- LinkedIn: https://linkedin.com/in/louis-peyrondet
- Portfolio: https://porococo.github.io/

---
Source: https://flows.cv/louispeyrondet
JSON Resume: https://flows.cv/louispeyrondet/resume.json
Last updated: 2026-04-10