SENIOR MACHINE LEARNING ENGINEER | Predictive Modeling Expert | Data Science Specialist

PREDICTIVE MODELING EXPERT

Time Series - Causality - AutoML - MLOps - Industrialization

---

For the past 10 years, I have supported companies in the design and industrialization of AI/ML solutions. My goal is to maximize the return on investment of your data science and artificial intelligence projects, from ideation to production.

---

- SERVICES -

Ideation | Exploration
I assist you in identifying the potential of predictive AI within your organization by evaluating both opportunities and associated risks. Together, we will analyze the return on investment (ROI) and its impact on your organization.

Analysis | Prototyping
In collaboration with you, I analyze the identified value streams and develop prototypes or MVPs (Minimum Viable Products) to validate concepts before moving to the next phase.

Design | Development | Industrialization
I design and implement robust technical solutions, as well as their infrastructure and cloud architecture, to ensure seamless integration into your decision-making processes.

Technical Lead | Support | Optimization
I assist you in the development, evolution, or optimization of your existing models and pipelines, ensuring their performance and alignment with your business objectives.
 

---

- FOCUS & SPECIALIZATION -

Predictive Modeling
End-to-end development and industrialization of predictive solutions for sales forecasting, demand forecasting, workload prediction, churn prediction, and more.

Time Series Expert
Prediction, classification, segmentation, anomalies, time-to-event analysis.

• Modeling: Statistical, Bayesian, Machine Learning, Deep Learning (RNN, CNN, TFT), Time LLMs
• Time Series Databases: TimeScaleDB, InfluxDB, TDEngine
• Tools: Nixtla, Statsmodels, GluonTS, PyTorch, SkTime, Darts, SkForecast, Scikit-learn, XGBoost, LightGBM, CatBoost, Stumpy, PyOD, ADTK
• Foundational Models: Moirai, Chronos, TimeGPT-1, TimesFM, TinyTimeMixer, TimeLlama, Moment
• Methods: Local, global, hierarchical, recursive, ensembles.

AutoML
Complete automation of ML pipelines: feature generation and selection, hyperparameter optimization, combinatorial ensembles, model selection.
Tools: Optuna, Nevergrad, HyperOpt.

Systematic ML
Extending AutoML to the entire production workflow, reformulated as a multi-objective optimization problem. Approaching Machine Learning as a search problem in the space of algorithms, features, and hyperparameters.

Causality
Inference and modeling of causal relationships from data, with "What If?" hypothesis testing linked to predictive models.
Tools: Tigramite, CausalImpact, CausalPy, CausalML, DoWhy, EconML, CausalNex.

Operations Research | Optimization
Modeling and solving optimization problems. Linear approaches (LP, MILP), genetic algorithms, reinforcement learning, black-box/gradient, swarms.
Tools: Gurobi, CPLEX, Google OR-Tools, Pyomo, Pulp, Scipy Optimize.

MLOps | Industrialization | Observability
Expertise in designing and implementing distributed, microservices-oriented ML infrastructures, including model monitoring and observability.
Tools: Docker, Kubernetes, Argo Workflows, NannyML.
 

---

- SOFT SKILLS -

Holistic Approach
Thanks to my dual academic background, I’ve learned to bridge technical challenges with strategic and organizational realities. This enables me to propose solutions that align with my clients' objectives.

Analytical Mindset
With my background in engineering and physics, I’ve developed a strong analytical mindset. I enjoy approaching problems methodically, finding clear solutions to technical challenges.

Strategic Perspective
My education at ESSEC and in political science allows me to understand the business, operational, and human aspects of projects. I focus on accurately identifying business needs and translating them into concrete solutions.

Teaching and Simplifying
Through my teaching experience at EPITECH and CELSA, I’ve learned how to explain complex concepts to diverse audiences. This enables me to adapt my communication to the technical level of my audience, whether they are experts or not.

Adaptability
As a consultant, I am accustomed to changing contexts. I quickly adapt to project needs without losing efficiency.

Resilience and Tenacity
Competitive sailing, particularly offshore racing, has taught me resilience in the face of unexpected challenges and adversity—qualities I apply to successfully completing the projects entrusted to me.