Guillaume Saës

PRESENTATION:

PhD. in Mathematics & Senior Data Scientist, I combine AI expertise with experience in data analysis and signal processing. I have developed advanced Machine Learning models, including LLMs like RaG models, in various sectors (health, transport, industry).

My approach combines scientific rigor and creativity to impute data, optimize complex processes, and design innovative models. I also popularize science through my YouTube channel (800,000 subscribers).

Skills: AI/ML (Deep Learning, LLM), time series, data imputation, visualization, optimization.

Technical Expertise: Python (Scikit-learn, TensorFlow, PyTorch), Matlab, Dataiku, CUDA, advanced statistics.

I transform ideas into concrete solutions, adapted to needs.

PROJECTS

- Development of RaG (Retrieval-Augmented Generation):

Applications in various sectors, notably health (CNAM) and others.

- Missing Data Imputation:

Development of Deep Learning algorithms to handle incomplete data.

- Time Series Forecast:

Participation in projects to predict railway traffic time series.

- Optimization:

Development of constrained optimization algorithms (MILP) for industrial projects (e.g., vehicle transport).

- Signal and Image Analysis Projects:

Multifractal analysis of physiological data from marathon runners and MEG

- Mathematics Research:

Thesis on multifractal analysis and characterization of signal regularity using wavelets.

- Scientific Popularization Projects, Science Trash:

Creation, writing, and production of videos for a broad audience (800,000 subscribers). Management of a team of 20 people to produce engaging content.

BACKGROUND:

- Data Scientist Quantmetry

- Mathematics Aggregation + Mathematics Thesis (Paris)

- Assistant in Probability and Statistics Department at Mons

- YouTuber (Science Trash 700k subscribers)