Apptainer vs Docker: HPC Parallel Workloads
In exascale-bound labs tackling multi-node MPI for fluid dynamics or climate sims on SLURM clusters, containers must integrate without friction.
Read on LinkedIn
Geospatial Data Scientist & Computational Hydrologist
I ship production geospatial systems: flood-risk models, climate data pipelines, and energy-infrastructure tools deployed across Europe, North Africa, and MENA. Python-first, HPC-ready, GIS-native.
Hydraulic engineer by training, geospatial engineer by practice.
I build end-to-end spatial workflows: ingest satellite data, run hydrological models, store in PostGIS/BigQuery, and serve via APIs or dashboards. I keep pipelines reproducible, tested, and documented.
Experience across research (INRAe), energy (ENGIE/CNR), catastrophe risk (REOR20), and public infrastructure (ANBT) keeps the work grounded in real constraints and operational delivery.
Built production 2D flood pipelines and a serverless hazard platform.
Geospatial data systems for hydropower operations and renewable planning.
Geospatial infrastructure for national water-resource planning and dam siting.
Tools and technologies used to build production geospatial systems
Vector/raster processing and GIS platforms.
Large-scale data manipulation and array computing.
Geospatial storage, indexing, and querying.
Scalable compute for geospatial workflows.
Satellite imagery analysis and deep learning.
Web services, mapping, and data visualization.
Core languages, version control, and delivery.
Selected work demonstrating technical depth and domain expertise
Geospatial AI pipeline for land-use and environmental monitoring in arid regions using Sentinel-2 and U-Net segmentation.
2nd place, Water Technologies Hackathon. Built a micro-hydroelectric optimization model in 48h for rural water/energy design with Python + EPANET.
Interactive flood-extent simulation with depth-damage curves, mitigation ROI, and scenario exploration.
Network-based agricultural supply chain analysis with vulnerability assessment, routing, and climate risk overlays.
Sharing knowledge on HPC and geospatial workflows
In exascale-bound labs tackling multi-node MPI for fluid dynamics or climate sims on SLURM clusters, containers must integrate without friction.
Read on LinkedInIn HPC clusters running distributed hydrological models, the choice between module environments and Apptainer containers determines reproducibility.
Read on LinkedInOpen to consulting, research collaboration, and full-time opportunities in geospatial data science, climate risk analytics, and sustainable infrastructure.