In this digital age, the LA POSTE Group has set itself the goal of becoming a data-driven company, placing data at the heart of its strategic priorities. As part of his role within the Mail and Parcels Service Branch, Antoine DE CHANTERAC, Project Director for traffic forecasting and econometric modeling, is exploring artificial intelligence techniques to enhance his current and future projects. He attended the “Introduction to Machine Learning and Deep Learning” training course offered by EURODECISION and shares his experience with us.

 

  • What motivated you to attend this training?

La Poste encourages its employees to train in machine learning. I had previously completed a 4-hour e-learning course covering the general challenges and ethical issues related to artificial intelligence. However, this was just an introduction, and I wanted to deepen my understanding since I currently manage a project where algorithmic intelligence plays a key role.

Two days are obviously not enough to become a machine learning expert, but the EURODECISION training provides a comprehensive overview of available techniques, including their respective advantages and drawbacks. It serves as an excellent starting point for those seeking to better understand the various technical challenges that may arise in such projects.

 

  • The training is based on fictional cases. Were you able to envision practical applications for your own work?

I believe illustrating concepts with examples is crucial, as the training content is quite dense, leaving little time to reflect on personal cases during presentations. In the intercompany session I attended, postal logistics was not among the case studies, but this did not bother me since my goal was to gain general knowledge.

The training days also included dedicated time for participant discussions, which was highly enriching. Engaging with professionals from diverse backgrounds allowed me to discover the challenges others face and to consider how machine learning could be applied in different contexts.

 

  • What key takeaways do you have from this training?

My goal was not to memorize everything but to gain a broad understanding of the available machine learning and deep learning techniques. I particularly noted that each project is unique, shaped by the company’s context, data, and the client’s expectations and needs.

Now, when faced with a new challenge, I know where to start and what to focus on. I took note of the many references provided and will consult them as needed.

 

  • What do you consider the strengths of this training? Would you recommend it to your peers?

Without hesitation, I would highlight the quality of the trainers. They are not just instructors or skilled mathematicians but true experts who regularly tackle real-world business projects and the questions they raise. They emphasize that model development must align with client needs and remind us that there is no one-size-fits-all solution—no ultimate algorithm exists.

I recommend this training to non-specialists involved in machine learning projects, such as project managers who need to collaborate with technical experts. The broad cultural foundation gained over these two days is a significant asset for navigating such projects more effectively.