Beyond the algorithm: what really makes AI-driven S&OP work.

Beyond the algorithm: what really makes AI-driven S&OP work ALL IN 2026, Montréal · Panel with Angelcare, Talan and TrueGradient On stage at ALL IN 2026 in Montréal, three perspectives that rarely share a platform came together around a real AI-driven S&OP project: the buyer, Stéphane Huot of Angelcare; the integrator, Jacques-Antoine Ollier of Talan; and the platform builder, Jasneet Kohli of TrueGradient. Their conclusion was blunt. Most AI transformations don’t stall on the algorithm. They stall on data nobody owns, a process that was never designed for it, and teams that were never truly brought along. In twenty candid minutes, the panel covers what has genuinely changed in the last eighteen months: forecast accuracy, it turns out, is not the headline. It’s AI’s new ability to explain a forecast in plain language and to reason across constraints and past decisions. They explain why pilots that try to “put AI in the supply chain” quietly die, while those anchored to one painful Monday-morning decision and a committed business number reach production. Stéphane shares what Angelcare underestimated: how much knowledge lives in people’s heads. The panel is also honest about where AI stops. It can’t create consensus, forecast true unknowns or fix a decision nobody owns. Watch to the end for the one piece of advice every supply chain leader starting this journey should hear before signing anything.