Freelancer profile translated to English.
Back to original languageAbout Glacien
Who I am
Data Scientist and ML Engineer with proven experience on critical data projects within major industrial groups: TotalEnergies, Hitachi Rail, L'Oréal, Thales, Sercel.
What I do
Machine Learning, Deep Learning & AI Integration
- Supervised and unsupervised predictive models (classification, regression, clustering, anomaly detection)
- Deep learning: CNN (computer vision, U-Net type segmentation), RNN/LSTM, Transformers
- NLP: information extraction, text classification, embeddings, LLM fine-tuning
- Uncertainty quantification (Bayesian, ensembling, calibration) key differentiator for industrial cases where reliability is paramount over raw accuracy
Data Engineering & MLOps
- Data pipelines (Python, SQL, Spark)
- Industrialization: Docker, CI/CD, orchestration (Airflow), experiment tracking (MLflow)
- Cloud: AWS, Azure, GCP
Hybrid Physics / ML Approach
Integrating domain knowledge into models (physics-informed ML)
Particularly relevant for sensor signals, industrial time series, geophysical data
What I'm looking for
Long-term contract missions (3-6 months, full-time) on high-stakes operational data science / ML challenges. I particularly enjoy contexts where data is imperfect, the technical domain is complex, and decisions are critical – that's where my industrial experience makes the difference.
French
Native or bilingual
English
Fluent
Can work on-site
Paris (up to 50km)
Experience
- HitachiRailData Scientist - Predictive Maintenance on Railway Signaling Systems Hitachi RailMECHANICAL ENGINEERINGSeptember 2025 - March 2026 (6 months)Paris, FranceDesign and industrialization of predictive maintenance models for a fleet of railway signaling systems (onboard assets, ground, and control centers) deployed internationally (Europe, Asia, Americas).Key achievements:Development of ML models (supervised and unsupervised) for early anomaly detection on sensor signals and time series, processing several hundred GB of IoT dataDesign of near real-time processing pipelines, from sensor collection to operational alertsReduction of false positives through continuous model optimization, replacing a preventive approach based on fixed thresholds / schedulesFacilitation of business workshops within a data/AI team of 5+ people, explaining results to operational signaling teamsValidated POC, currently being deployed on the operational fleetImpact: Transition from a calendar-based preventive maintenance logic to a predictive logic in a critical railway safety domain, direct contribution to the operational availability of signaling systems on an international scale.Stack: Python, SQL, time series, ML/DL, Git, Power BI, Jupyter
- LorealData Scientist - ESG & Decarbonization Reporting L'OréalENVIRONMENTALJune 2024 - July 2025 (1 year and 1 month)Design and industrialization of a consolidated carbon reporting system to guide the group's decarbonization strategy for France, replacing a time-consuming manual Excel process.Key achievements:
- Construction of a standardized ESG data model, consolidating heterogeneous multi-entity sources within the French scope
- Full development of the Power BI dashboard (DAX, dynamic measures, automation of monthly refreshes aligned with CSR reporting cycle)
- Python scripts (Pandas, NumPy) to standardize and improve the reliability of upstream processing
- Anomaly detection, trend breaks, and inconsistencies in carbon indicators
- Cross-functional coordination with CSR teams, data owners, and management; training of end-users
- Solution adopted by 10-30 regular users (CSR teams, management, zone managers)
Impact:transformation of a manual ESG reporting process (Excel + emails, several days per cycle) into a centralized, automated, and reliable BI solution. Significant acceleration of strategic decision-making on the French group's carbon trajectory and improved monitoring of decarbonization commitments.Stack:Python (Pandas, NumPy), Power BI, DAX, M, SQL, SAS - ThalèsData / ML Engineer - Automation & Infrastructure Supervision ThalesFILM AND AVJuly 2022 - April 2024 (1 year and 9 months)Vélizy-Villacoublay, FranceDevelopment of a Python automation solution for the secure deployment and supervision of a fleet of collaborative equipment across most Thales France sites.Key achievements:
- Development of an on-premise configuration tool via REST API, capable of configuring several hundred pieces of equipment according to Thales security standards
- Reduction of unit configuration time from several hours to a few minutes per equipment, directly benefiting IT and network teams
- Implementation of an automated system for real-time collection and monitoring of usage data (previously non-existent fleet visibility)
- Collaborative development of a firmware version adapted to the group's specific requirements, ensuring compliance with internal security standards
- CI/CD integration and deployment industrialization (Git, GitLab, Bash)
- Multi-site coordination with IT, network, and support teams across the entire France deployment
Impact:homogeneous deployment across France, significant reduction in configuration errors and intervention times, guaranteed security compliance across the entire fleet, and increased stability thanks to automated pipelines.Stack: Python, REST API, CI/CD, Git/GitLab, Bash, DevOps
Recommendations
Be the first to recommend Glacien
Help this freelancer shine by sharing your experience working together.
These freelancer profiles also match your criteria
Agatha Frydrych
Backend Java Software Engineer
4.7
(3)
2
Baptiste Duhen
Fullstack developer
4.6
(4)
5
Amed Hamou
Senior Lead Developer
4
(2)
7
Audrey Champion
Web developer
4.3
(3)
4
Education
- Deep Learning Specialization Bootcampdatascientest.com2021•Compétences principales : Formation pluridisciplinaire combinant l’analyse de données, la modélisation mathématique et l’IA (apprentissage automatique et apprentissage profond). •Domaines d’expertise : Développement d’algorithmes d’apprentissage automatique, analyse statistique, exploration de données et gestion de bases de données. •Applications : Interprétation des données, modélisation probabiliste, inférence statistique, visualisation des tendances, nettoyage et transformation des données, et gestion de bases de données SQL et NoSQL.
- Data Science EngineerPolytech Sorbonne2020•Compétences techniques : Maîtrise de Python, R, MATLAB et SQL pour l’analyse de données, la modélisation et la gestion de bases de données. •Expertise en apprentissage automatique : Application de l’apprentissage automatique et de l’apprentissage profond à l’interprétation de données, y compris l’analyse de séries temporelles et le traitement de signaux. •Domaine d’études : Spécialisation en géophysique computationnelle et analyse de données, combinant la modélisation mathématique, l’apprentissage automatique et la programmation.