Yannick Raharijaona

Levio consulting to an in-house AI branch called "Menya Solution" which generally does research and development in Artificial Intelligence ranging from classic artificial intelligence (fuzzy logic, expert system) to machine learning (reinforcement learning, supervised learning, unsupervised learning). I had the opportunity to work on a computer vision project that detects trouble makers in a crowd, my contributions to the project are: – Assuring the quality of the data by checking if the bounding box annotes are in place or not and make sure that they have the right labels. – Coding the data pipeline to have smooth and high-quality data transmission. – Challenging the FairMOT approach by providing critical thinking, experimenting with your thoughts and validating them by doing unit tests. Technology uses: Pytorch, Kafka, Docker.

To conclude, my master's degree in data science, I spent a 4-month internship at Intact Insurance. I had the privilege of working on their internal framework, the main tasks of which I was entrusted with are: – Extend their data visualization libraries by adding a feature allowing to compare the performance of two models generically depending on the month. – Optimizing their feature store by fixing bugs due to various reasons. As well as documenting, the entire feature store. Technology uses: Numpy, Pandas, git, linux.