Thomas Jaillon

- Developed time series forecasting solutions within smile R&D, utilizing cloud cluster data. - Implemented various approaches including Statistical Models (SARIMA), Machine and Deep Learning (VAR, Xgboost, LSTM), and Hybrid solutions (LSTM-VAR). - Implemented a Docker stack using Prometheus (TSDB) x Grafana (dashboard) for monitoring cloud cluster metrics. - Created a new metric and visualization for evaluating time series forecasts. - Created a Python library dedicated to time series forecasting.

- Within the EDF R&D intellectual property office, I designed and developed solutions for managing patent portfolios from scratch. - Gained insights into the intellectual property domain, specifically patent processes and associated challenges, while working within EDF R&D. - Created strategic data visualization tools using Dash in Python, enabling patent portfolio exploration and predicting patent costs.