24 au 26 février, 2027
Montréal, Canada
Forecasting without training: time-series foundation models
Foundation models did not stop at text. Pretrained on billions of points, time series models now forecast series they have never seen, with no training at all. We put one in production and retired the models we trained per metric. You will learn how zero-shot forecasting works, when it beats training on your own data, and what it costs you.
Denis Roussel
Centreon
Denis Roussel builds the AI inside Centreon's monitoring and log management products: forecasting, anomaly detection, and agentic log analysis. He runs all of it in production. He has specialized in time series for the past decade, and now spends as much time on release pipelines, model licensing and developer tooling as on models themselves.
