Consultant in Data Science and Optimization
2016 - PRESENTFreelance- Helped a large public institution detect fraud using semi-supervised learning and graph algorithms, using Python and Spark.
- Worked on a large optimization model that generates schedules of nurses in hospitals automatically, using Python and Gurobi.
- Used machine learning to generate weekly forecasts of oil demand in various parts of the world for a tanker company.
- Worked for a bank as a Python solution architect to transform a monolithic calculation engine to a full-fledged distributed application.
- Gave training sessions about data science, forecasting, Python, and R.
- Developed a mixed-integer programming (MIP) model that optimizes the production planning for a large biotech company.
Technologies: Celery, Redis, Jenkins, Docker, Spark, SQL, Gurobi, Git, Java, RStudio Shiny, R, Jupyter, PythonTech Co-founder, Director
2015 - 2016Reveal- Developed an algorithm used by head hunters that match people and jobs using machine learning.
- Founded the company together with Founders, a venture capital firm based in Denmark.
- Used the same technologies as we developed for head hunting to provide insights about the job market.
Technologies: Natural Language Processing (NLP), PythonAdvanced Analytics Team Lead
2013 - 2015Maersk Line- Started major projects and developed models together with my team.
- Initiated various projects involving weekly forecasts with a geographical dimension, models to analyze costs, etc.
- Worked on each stage of the project from the initial phase of the problem formulation to the development of models and the final presentation of the solution.
Technologies: Java, SQL, ROptimization Specialist
2010 - 2013Maersk Line- Worked on a very large multi-commodity flow model.
- Developed a web interface that enabled end users to easily interact with the model.
- Helped reduce the consumption of energy significantly via the model that we developed.
Technologies: Computational Infrastructure for Operations Research (COIN-OR), C++, Git, JavaEngineer in Applied Mathematics - Optimization and Modeling
2006 - 2010Descartes- Designed and implemented algorithms in the transport optimization area. The work consists in solving real-life vehicle routing problems (VRP) and their variants.
- Split very large problems in smaller ones to reduce the complexity of problems such as client clustering and mono-route optimization.
- Developed technology for waste management to facilitate the transport of containers filled with industrial waste to depots.
Technologies: GNU, C++Consultant
2004 - 2006n-Side- Worked on a non-linear optimization model, able to choose the best combination of raw material to produce steel at the lowest cost and satisfy quality requirements.
- Worked on various forecast models to predict the demand of daily newspapers and to predict the rise in the water level in Belgian canals.
- Helped a large chemical company optimize their supply chain and their distribution network.
Technologies: CPLEX, Java