Steven Poivre

Steven Poivre is CEO and co-founder of Praedixa, an AI-native platform dedicated to operational forecasting for multi-site restaurant networks and franchises.

A Data Scientist specialized in time series and demand forecasting, he designs the models that anticipate cover volumes, product-level sales and labor needs at D+7 and D+30. The Praedixa approach combines POS history, school calendars, local events and exogenous signals (weather, holidays, neighborhood dynamics) to transform raw restaurant data into concrete operational tradeoffs.

His technical conviction: in foodservice, much of the waste, stockouts and overstaffing is not a business inevitability — it is a signal-capture problem. When a demand model has access to the right data and is wired into field decisions (purchasing, scheduling, production), the margin gain far exceeds software cost. That premise structures the Praedixa product.

Today, Praedixa supports restaurant networks that want to steer operations with the rigor of reliable data, without sacrificing the field autonomy of their managers. The platform sits above existing tools (POS, scheduling, inventory) to enrich them with a usable demand signal, rather than replacing them.

Steven writes about operational forecasting in foodservice, the antipatterns of "by-habit" scheduling, the integration of AI models into business workflows, and multi-site network pilot approaches.