Summary
Computer science PhD based near Morlaix, with a background in data science and a current focus on data engineering and software architecture for weather forecasting and cartography-based applications.
My career bridges research, engineering, and product development, with strong expertise in data-intensive systems, geospatial technologies, and applied machine learning.
Through academic and industrial projects, I have led and contributed to end-to-end data platforms, from proof of concept to production, industrialization, and large-scale operation.
Experiences
- Web application & geospatial stack
- Leading the development of a yachting / shipping web application
- Building interactive web mapping features with Leaflet
- Implementing geospatial data visualization for marine weather use cases
- Data & database architecture
- Designing and maintaining the PostgreSQL / PostGIS data layer
- Modeling spatial and time-series weather data
- Technical leadership & code quality
- Defining and enforcing engineering best practices
- Ensuring clean architecture, modular design, and feature isolation
- Maintaining high standards for testability, versioning, and documentation
- Team growth & delivery
- Conducting code reviews and defining testing strategies
- Participating in technical recruitment
- DevOps & reliability
- Contributing to CI/CD pipelines and deployment workflows
- Improving platform reliability, monitoring, and operational practices
Python TypeScript PostgreSQL/PostGIS
GraphQL Svelte/SvelteKit Docker/Swarm Hasura DuckDB
- Analysis of scalability challenges on the IP traffic supervision system (Netflow) from a SD-WAN product
- Leveraging the existing, in-production OLTP system (custom aggregation, time-partitioning)
- Deep study of alternatives OLAP systems
- Implementation of testing/simulation tools, generation of massive realistic network data
- Upstreamed contributions to
scapyNetflow module
- Upstreamed contributions to
- Architecture and implementation of the alerts/monitoring database of a network management solution (NMS)
- DBMS optimization (Index tailoring, request profiling)
- Automation of embedded database integration pipelines for routers (traffic classification)
Python asyncio Galera Cassandra
Go DuckDB PostgreSQL MariaDB pola.rs scapy Gitlab CI/CD
- Regression, Classification, Forecasting on various industrial topics
- agronomy
- logistics
- Natural Language Processing (NLP)
- topic segmentation
- sentiment analysis
- Image analysis (Computer Vision, anomaly detection, CNN)
- production monitoring
- quality control
- Deployment and industrialization of Data Science projects (MLOps)
- Report writing and project presentation to stakeholders
- Audit of business operations, coordination with IT/Datalabs, writing of specifications
- Writing and conducting pedagogical trainings
- “Lean 6-sigma black belt” level
- Qualiopi certified on first session
- Mentored:
- 4 internships (BsC and MSc level)
- 3 phd candidates in a summer school program
Python PySpark PyTorch gensim Keras OpenCV
R PostgreSQL SnowFlake Docker Anaconda
- Machine learning applied to structure-activity relationships (QSAR)
- Development of Feature-Learning algorithms for for molecular subgraphs
- Exploration of correlations between topological and macroscopic models
- Contributed to graph isomorphism problem, library published in MIT
Worked on side subjects as Junior Data-Scientist.
Python PySpark PyTorch gensim R Docker
Education
Conducted with success and large autonomy a 3 years data science project involving:
- academics: IRISA (Expression team) & LMBA
- industrials: Avril group & See-d (small scale research lab)
Manuscript available on TEL
Projects
- canonization algorithm working on fully labeled graphs (vertices and edges)
- provides for any graph a tree representant of its isomorphism class
- well suited for low-connected graphs (e.g. molecules)
- Has been used to provide real-time access to a pricing model (retail) running in R
- Key point is to pipe the incoming HTTP request to a pool of interpreters kept opened
- < 20ms of additional latency.
OSS Contributions
Some open-sources projects I enjoyed contributing for:
Publications
Skills
Technical Stack
Python asyncio pola.rs pandas scikit HuggingFace Spark PyTorch
R SQL Go Node.js
bash Docker Gitlab CI/CD MLFlow AirFlow Prometheus Grafana
DBMS
-
OLTP
PostgreSQLMariaDBGalera -
OLAP
ClickHousepolarsDuckDBSnowFlake -
NoSQL
CassandraMongoDBRedis
Data Science
- Regression, Classification, Forecasting
- Feature Selection
Data Engineering
- DBMS optimization (Indexing/Sharding, Profiling, High-Availability)
- ETL (
AirFlow,Kafka,Spark)
Machine/Deep Learning
- Transformers
- Auto-Encoders
- Generative Adversarial Networks
- Convolutional Neural Networks
- Q-Learning
Algorithmics
- Graph Theory
- Algorithms Complexity
- Distributed Computing
- Asynchronous Programming
Natural Language Processing (NLP)
- Topic segmentation
- Document model
- Semantic vectorization models (word2Vec, GloVe)
Computer Vision (CV)
- Object Detection/Segmentation
- Pattern matching
- Feature Extraction
Communication
- Reports writing & presentations
- Trainings writing & animation