Jérôme Gurhem

August 2021 - Present ANEO, Boulogne-Billancourt, France Product, Engineering & Research Lead, HPC/Cloud

ArmoniK — Open-Source HTC Orchestrator August 2021 - Present

• Engineering
  • Led the full lifecycle of ArmoniK, an open-source task-based high-throughput computing (HTC) framework for orchestrating large-scale, fault-tolerant workloads across distributed and elastic infrastructures, processing tens of millions of tasks per month in production for major financial institutions
  • Architected and implemented its core on Kubernetes and Terraform, driving its evolution from initial design to production-grade deployments scaling to 12,000+ concurrent workers on a single instance within a single cloud region in hybrid multi-cloud environments
• Product
  • Owns ArmoniK's multi-year strategic product roadmap and oversees a portfolio of ArmoniK projects representing €700K–€1.2M in annual revenue, defining technical direction, feature prioritization, and delivering major releases on a 6-month cadence in alignment with client needs and market trends
  • Manages ArmoniK's budget and FTE resource allocation, making investment decisions across R&D, engineering, and support — spanning headcount planning, contractor engagement, and cost optimization
• Research & R&D
  • Led R&D for ArmoniK, contributing to scientific publications, and leading Aneo's involvement in the FPT4 project (successor to InterFLOP)
  • Co-supervised a PhD thesis on the formal modeling, verification, and optimization of ArmoniK, bridging academic research and industrial application
• People
  • Managed and mentored a cross-functional team of ~10 PhD researchers, engineers, and interns, fostering technical growth and delivering high-quality software components
• Operations
  • Scaled and structured ArmoniK's customer support operation for Crédit Agricole CIB and established it from the ground up for Natixis — across 30+ deployments (dev, integration, pre-prod, production) for 6 client teams; defined and delivered Level 3 support services, built issue-tracking and client follow-up workflows, and coordinated bug triage and corrective patches
  • Actively participated in pre-sales engagements, contributing to technical proposals, architecture workshops with prospective clients, and commercial positioning of ArmoniK for regulated industries (finance, insurance)
  • Created and launched ArmoniK's training offering, designing the curriculum and delivering 5+ multi-day sessions (presentations, hands-on labs, hackathons) to 5–25 participants per session, both internally and to external clients as a commercial service

Agentic CFD Simulation Platform March 2026 – May 2026

• Co-designed the SaaS platform layer for an Agentic CFD Simulation Platform combining a GPU-native CFD solver (AWS ParallelCluster/PCS, NVIDIA H100 GPUs) and a vector database (PostgreSQL with pgvector); designed the platform architecture including identity and access management, usage tracking, platform governance, web-based user interface, and future Bring-Your-Own-Cloud (BYOC) deployment model with usage-based pricing
• Co-designed the agentic AI orchestration architecture using LLM-based agents (AWS Bedrock with Mistral models) and multi-agent orchestration protocols (MCP, A2A), enabling natural language-driven simulation workflows with human-in-the-loop control across the full simulation lifecycle (configuration, execution, monitoring, analysis); integrated observability via Langfuse and SigNoz
• Coordinated cross-team delivery across a 7-person team spanning the SaaS platform, the GPU/HPC infrastructure team, and the AI agent development team

People & Organization August 2021 - Present

• Acted as Manager Référent, supporting consultants' career development through roadmap planning, feedback collection, and alignment with Aneo's culture and HR processes, complemented by hands-on technical coaching and long-term mentoring of junior engineers, PhD researchers, and interns
• Actively involved in strategic recruitment: defined job descriptions, conducted ~20 technical interviews per year, and contributed to onboarding and long-term hiring decisions

March 2021 - July 2021 CNRS USR3441 Maison de la simulation, Saclay, France Postdoctoral researcher

• Contributed to the development and parallelization of a Graph Convolutional Neural Network (GCNN) training and inference library, optimizing for large-scale graph data processing
• Ported a hybrid MPI + OpenMP implementation of Sparse Matrix-Vector Multiplication (SpMV) to the A64FX architecture, leveraging SIMD vectorization for performance gains
• Conducted performance benchmarking and scalability analysis on the Fugaku supercomputer, evaluating efficiency across thousands of cores

March 2018 - March 2021 CNRS USR3441 Maison de la simulation, Saclay, France Ph.D. Student

• Conducted a comprehensive survey and classification of task-based programming paradigms for high-performance computing applications
• Designed, implemented, and deployed task-parallel scientific applications on large-scale supercomputing platforms
• Performed experimental evaluations of block-based direct solvers for dense linear systems, sparse matrix-vector multiplication (SpMV), and Kirchhoff pre-stack depth migration in seismic imaging workflows
• Partially supported by TOTAL SA

March 2017 - August 2017 CNRS USR3441 Maison de la simulation, Saclay, France Intern — Parallel Programming

• Developed and tested parallel applications using YML and XMP programming models on high-performance computing platforms
• Conducted performance evaluation and scalability analysis across multiple scientific workloads
• Designed and prototyped a new parallel implementation of Kirchhoff pre-stack depth migration, targeting improved efficiency and scalability on supercomputers

May 2016 - July 2016 Norman Info, Arras, France Intern — Jira Platform Migration

• Migrated the company's project tracking platform across multiple major Jira versions, designing a 3-stage intermediate migration path to safely bridge a large version gap
• Organized and delivered team training sessions to onboard staff on the migrated platform
• Applied Scrum to plan and coordinate the migration timeline, gathering feedback to refine configuration post-migration

August 2015 French Tax Office, Saint-Maur-des-Fossés, France Intern — Administrative Operations

• Understanding the sorting system of a big company
• Managing time efficiently to do all the tasks
• Helping others to be faster in their work

December 2022 - Present Université de Versailles Saint Quentin - High Performance Master, Guyancourt, France

• Classes: Numerical calculus
• Topics: Iterative methods to solve linear systems, band matrices, Lapack, BLAS
• 12 hours of practical courses.

January 2022 - May 2025 Institut des Sciences et Techniques des Yvelines, Vélizy, France

• Classes: Advanced Operating Systems
• Topics: Elf format, process scheduling, threading, virtual memory
• Student project: implementing a Shell in C
• Lecture on "Changes in application development brought by Cloud Computing"
• 2 hours lecture + 40 hours of practical courses.

March 2018 - June 2018 Institut des Sciences et Techniques des Yvelines, Vélizy, France

• Classes: Operating Systems
• Topics: Unix basic commands, bash scripts, C processes
• 60 hours of practical courses.

March 2018 - March 2021 Ph.D. in Computer Science

Thesis title: Distributed and Parallel Programming Paradigms Using Graphs of Tasks for Post-Petascale Supercomputers
Advisor: Pr. Serge G. Petiton
University of Lille (Lille, France)

September 2014 - September 2017 Polytech'Lille, Computer Science and Statistics

Master in Computer Science and Applied Mathematics
University of Lille (Lille, France)
Degree project: Distributed and parallel multi-level programming on supercomputers using YML+XMP, a development and execution environment for parallel and distributed applications based on a graph of PGAS components managed at runtime

September 2016 - February 2017 School of Computer Engineering

Master in Informatics Engineering - Erasmus
Technical University of Madrid (Boadilla del Monte, Spain)

September 2012 - July 2014 Undergraduate Studies

Preparatory Classes in Mathematics and Physics: Specific training for competitive entry examinations for the French "Grandes Ecoles"
Lycée Clémenceau (Reims, France)

September 2009 - July 2012 Scientific Baccalauréat

Major: Mathematics
Lycée Samuel Beckett (La Ferté-sous-Jouarre, France)

December 2021 (expiration: December 2023) Google Cloud Certified - Professional Cloud Architect

25 January - 8 February 2021 (1 d/week) [ONLINE] Modern Scientific C++ @ MdlS/Idris

November - December 2020 [MOOC] Deep Learning (CNAM)

7 July - 9 September 2020 (2 h/week) [ONLINE] Kokkos Lecture Series

18 December 2019 Introduction to machine learning in Python with Scikit-learn

4 - 5 July 2019 Introduction to Parallel Programming with HPX

12 - 14 June 2019 CEA-Riken Summer School - Mastering the Arm HPC ecosystem

23 - 24 April 2018 Parallel file systems and parallel IO libraries