Mistral says workflows now in production at ASML, ABANCA and CMA‑CGM

Enterprise AI is moving past chatbots and into the boring, expensive stuff that actually runs companies. That is the point of Mistral’s new Workflows product — not to write prettier text, but to keep AI-driven business processes from breaking in production. On April 29, Mistral put Workflows into public preview and said customers including ASML, ABANCA, CMA‑CGM, France Travail, La Banque Postale, and Moeve are already using it for live business processes at scale. (mistral.ai) ### What is Mistral actually launching? Workflows is an orchestration layer inside Mistral Studio. Basically, it lets a company stitch together models, tools, rules, and ordinary code into a repeatable process that can pause, retry, log every step, and hand off to a human when needed. That sounds unglamorous, but it is the difference between a demo and something a bank or shipping group will trust. (mistral.ai)? Most enterprise AI failures do not come from the model forgetting a fact. They come from the workflow around the model being fragile — a timeout, a bad tool call, a missing approval, an unlogged exception. Mistral is pitching durability, observability, and fault tolerance as the fix. In plain English, the company is saying: your AI process should survive the messy parts of real operations, not just work in a notebook. (mistral.ai) ### Who is already using it? Mistral named six customers in the launch post: ASML, ABANCA, CMA‑CGM, France Travail, La Banque Postale, and Moeve. Those are not random logos. They span semiconductors, banking, shipping, public services, and energy — exactly the kinds of sectors where uptime, audit trails, and data controls matter more than a flashy consumer app. (mistral.ai) ### What are those companie(mistral.ai)ational. It points to freight release flows, customer-data checks, and specification reviews — the kind of tasks that mix documents, rules, approvals, and human signoff. On its solutions pages, Mistral also highlights CMA‑CGM’s internal assistant, MAIA, as serving a workforce spread across 160 countries and more than 155,000 employees. (the-decoder.com)orchestration-with-workflows/)) ### Why does “millions of daily executions” matter? Because that is the strongest signal in the whole announcement. AI vendors say “enterprise” all the time, but volume is what separates a trial from infrastructure. Mistral is telling buyers that these are not side experiments — they are repeated, production-grade processes running often enough to matter to the business. The catch(the-decoder.com)gregated. (mistral.ai) ### Why are ASML and Europe so central here? ASML is not just a customer name-drop. It announced a strategic partnership with Mistral in September 2025, tied to a long-term collaboration around AI across ASML’s products, R&D, and operations. ASML also led Mistral’s €1.7 billion fundraising round, which turned the relationship into a broader bet on European AI infrastructure, not just model procurement. (asml.com)rategic-partnership)) ### Why does sovereignty keep coming up? Because Mistral’s sales pitch is not only about capability. It is also about control — open models, deployment in a customer’s own environment or private cloud, and keeping sensitive enterprise data inside regulated boundaries. That lands especially well with European banks, public agencies, and industrial groups that do not want core workflows tied to a black-box U.S. SaaS stack. (mistral.ai) ### So what changed? Mistral used to be judged mostly as a model company. Now it is trying to move one layer up the stack and own the runtime where enterprise AI work actually gets done. If that works, the company is not just selling intelligence — it is selling the system of record for AI operations. (mistral.ai)