Scientific data logistics. Exascale workflows.
Computer scientist researching high-performance computing and cross-facility scientific workflows for radio astronomy.
Research
Doctoral thesis
Collaborative System of Systems for Scientific Data Logistics
Scientific instruments produce more data than any single site can receive, process, or keep. Turning those flows into science-ready products means running chains that cross instruments, supercomputers, storage platforms, and research networks — each under a different administration and a different law.
That is a governance problem before it is a bandwidth problem. The thesis models this data logistics, optimises it against competing criteria, specifies the metadata it needs, and exposes it to scientists without hiding the requirements that keep results secure and reproducible.
- ~700 PB — of SKA science-ready data archived every year
- 3 — French national supercomputing centres, all subject to PPST and ZRR
- ~7.9× — measured DDF Pipeline speedup on 24 nodes — 68.8 h down to 8.7 h
- Radio astronomy — Grounded in the DDF Pipeline for LOFAR, an SKA precursor, running across the French national supercomputers and European research storage.
Watch & listen
- Interview: At the Heart of Data Logistics for Radio Astronomy — An ECLAT interview about Mathis’s path, doctoral research, and multidisciplinary work.
- Webinar: HPC Application Services for Radio Astronomy — A webinar on data logistics, intensive imaging, and large-scale workflow orchestration.
- Event: International Hackathon for Radio Astronomy — Highlights from a collaborative ECLAT research event in Rennes.
- Demo: MCP Server Exposing the Structured Knowledge Map of the DDF Pipeline Software — Demo of the MCP server exposing the software ontology for DDF Pipeline.
Latest entries
- Exascale Astronomy vs. Cybersecurity: Unlocking French Supercomputers for the Radio Astronomy Community — How a service-based access model can unlock French national supercomputers for radio astronomy without compromising cybersecurity.
- Predicting for Sizing: Teaching Supercomputers Their Own Execution Footprint — A continuous-learning predictive model that estimates the execution time, memory, and energy of scientific applications on supercomputers — so researchers can size their allocations instead of guessing.
- Interview: At the Heart of Data Logistics for Radio Astronomy — In this ECLAT interview, Mathis Certenais talks about his path from an engineering school in Rennes to doctoral research on data logistics — and why multidisciplinarity is key to tomorrow’s radio astronomy.
- YoungPEx: The New Generation of Digital for Exascale — Inside YoungPEx, the NumPEx working group that connects young researchers across the French exascale programme — its first General Assembly in Toulouse, its workshops on the environmental impact of HPC and on inequalities in academia, and what comes next.
- How I Created an AI Lab to Launch My Open-Source Virtual Running Coach — In this side-project blog, Mathis Certenais explains how a knee injury and a folder of GPS files became an open-source virtual running coach — and why the real product is the method behind it: one orchestrator, eight specialist agents, and a test suite that refuses to let a number lie.
- Make Scientific Software Speak: An MCP Server That Explains Its Own Parameters — Think of your scientific software documentation as an interactive wiki. An MCP server that transforms hundreds of complex configuration settings into natural-language answers for researchers and engineers in the community.
- International Hackathon for Radio Astronomy — A research event bringing together computer scientists and astrophysicists from France, England, and South Africa to address the shared technical challenges of radio astronomical data.
- Webinar: HPC Application Services for Radio Astronomy — This webinar uses radio astronomy to demonstrate how HPC Application Services can manage scientific data, accelerate imaging, and support next-generation instruments such as the SKA through architectures and tools for cross-facility workflows.
- From Supercomputers to Tutorials: Strengthening Our Community Through Shared Expertise — A technical research event bringing together computer scientists and astrophysicists to tackle challenges of radio astronomy.
In short
- Doctoral Research — Specializing in workflow data management, with a focus on governance and cybersecurity in supercomputing environments
- Published Work — Author of a full paper titled “Enabling Radio Astronomy to Use French Supercomputers,“ accepted for the Supercomputing Conference
- Scientific Collaboration — Working for the NumPEx "Digital for Exascale" research program and the ECLAT joint laboratory to co-design future radio astronomy workflows