Service - Data Modeling

Machine learning
Train machine learning (ML) models on any type of data from very simple and explainable models to more complex and performant models.
Deep learning
Train deep learning models (neural networks, RNN, CNN...) on images, text, mixed data. Implement custom architecture depending on the needs of the project.
Model monitoring
Monitor the models during training and in production. Ensure that the model is performing reliably and without errors weeks or months after training. Put in place alerts when the model stops performing at the required level.
Automated workflows
Create scripts to automatically retrain the model when new data are available or when performance are degrading.
Model hosting
Making the model available through an API internally to other people in the company or publicly available online. Training the model in a way that the preprocessing can also be performed by the API.

I trained hundreds of machine learning models on a dataset of tens of thousands of job offers. I performed experiment tracking for all these models allowing me to compare and select the best one. Once selected I hosted this model locally and online on AWS using an API.
I created a website to generate a preprocessing and modeling pipeline. The user can select from a combination of thousands of preprocessing steps and compared different models and their performances.

Data modeling