Two Mines Saint-Étienne Alumni Launch Odeai, a Very Short-Term Weather Forecasting AI

Paul Bonnet-Eymard and Thibaut Cassard have created an innovative tool for accurate weather forecasts within 3 hours.

Generic image of a weather forecast interface on a tablet.
AI

Generic image of a weather forecast interface on a tablet.

Engineers Paul Bonnet-Eymard and Thibaut Cassard, alumni of Mines Saint-Étienne (class of 2016), have founded Odeai, a company offering ultra-short-term weather forecasts updated every ten minutes.

After completing the Ingénieur Civil des Mines (ICM) program at Mines Saint-Étienne, Paul Bonnet-Eymard and Thibaut Cassard combined their expertise to create Odeai. The company develops weather forecast models with a 3-hour horizon, updated every ten minutes, integrating physics and artificial intelligence for enhanced accuracy.
Thibaut Cassard, a data scientist specializing in meteorology, previously designed forecasting models at Plume Labs and AccuWeather. Paul Bonnet-Eymard gained experience in marketing and business development at Air Liquide. Their solution, initially focused on rain prediction, targets sectors such as viticulture, major sporting events, civil security, and water network management.
Odeai aims to overcome the limitations of current tools by providing faster and more precise forecasts. "Weather, even in the very short term, influences very concrete decisions," explain the founders, citing the example of a vineyard treatment that needs reapplication if washed away by rain. The company plans to expand its forecasts to other variables like wind and sunshine, and to cover the energy sector.
Launched in the United States, Odeai will be available in France and Europe by the end of the year. The company aims to gradually extend the forecast horizon to 24 hours by the end of 2027.
Their time at Mines Saint-Étienne provided them with a strong scientific and technical foundation, as well as essential versatility to understand both technical and industrial challenges. They emphasize the role of AI in rapidly prototyping ideas, while stressing the importance of clearly defining the problem to be solved, advising to "often work on paper first before using an LLM".
Based on information from the official source: École nationale supérieure des mines de Saint-Étienne (02/11/2026)