Jean A.

- Setting up an AI product from the visual defect detection models developed by the research team

- Implementing the backend of the product, including the training and inference steps, image pre-processing and post-processing, the product API, communication with the robot driver, the implementation of the SQL database, etc.

- Develop and deploy ML/DL models for innovative projects of Amadeus' internal incubator.

- Design and build a production-ready and scalable ML architecture (training, inference, monitoring and versioning architecture, coding standards, MLOps flow with the Ops team).

- Use cases mainly NLP (deduplication of suppliers and products, automatic taxonomy, product clustering...) using standard approaches and transformer-based approaches (SBERT etc.)

- Supervising students following the 9-month data science training provided by Openclassrooms

- Project-based training aimed at helping students acquire the following skills.

▪ mastery of Python, Pandas, Numpy, matplotlib, scikit learn and Keras.

▪ cleaning, exploration, analysis and visualization of data, clustering, segmentation, feature extraction, machine learning and deep learning algorithms, explainability, etc