1,300 Restaurants, Six Brands, One AI Roadmap

"Five years ago, my challenge was to convince people AI could do amazing things. Now, my challenge is not to say no to everyone." Sophie Buresi is Group Head of AI at QSRP, the Zug-based quick-service restaurant platform that operates Burger King, Quick, O'Tacos, Nordsee, Dunkin and Chopstix across roughly 1,300 restaurants in multiple European markets. She joined into a brand-new role with a two-part mandate: use AI to improve EBITDA, and keep HQ lean as the group grows.  In this conversation with Conor Sheridan she walks through what the real job of a Head of AI actually looks like today. Why the biggest unlock is almost never AI itself (it's technical harmonisation and data foundations). How she decides which AI use cases to invest in across six brands and multiple geographies, using drive-through voice ordering as the worked example (live in Belgium and the Netherlands, deliberately paused in Italy because the unit economics don't work there). How she sells a use case to a franchisee who owns their own P&L, with payback measured in months not years. Why hardware-heavy deployments (kiosks, scales) are owned by the ops team, not Group AI. The two parallel HQ efficiency streams she's running: AI for All for individual efficiency, and process transformation for department-level change. Plus a quick-fire round on the Monday-morning metric Sophie actually looks at first, the AI category she thinks is overrated in QSR today, build vs buy, and the one piece of advice she'd give a CTO handed a brand-new AI mandate. KEY TAKEAWAYS The business case is local. Value per lane moves with the brand, the market and the store. You have to price AI use cases market by market, and keep measuring after launch — the use case that pays today can stop paying tomorrow, and you need to be ready to switch it off. The operator decides, always. Pilot in company-owned restaurants, pick franchisee lookalike locations (not just flagship city centres), and show payback in months not years. If a use case adds complexity, restaurant teams won't use it. The usage data will tell you straight away. AI is the cherry on the cake. The real work is harmonising systems, centralising data and fixing processes. Most of what it takes to deploy an AI use case is actually simplifying a process, not adding an agent. The Head of AI job has flipped. Five years ago it was convincing people AI could do amazing things. Today it is saying no to everyone until the data foundations, architecture and processes are in place that actually let AI pay. Operators have to be at the table from day one. Restaurant managers and district managers spot the operational blockers that no pilot framework will. Group AI designs the use case with local ops, not for them. USEFUL LINKS & RESOURCES Sophie Buresi on LinkedIn: https://www.linkedin.com/in/sophieburesigallay/ CONNECT WITH THE SHOW All What's Cooking episodes: https://www.nory.ai/podcasts Nory: https://www.nory.ai/ Nory blog: https://www.nory.ai/blog Nory on LinkedIn: https://www.linkedin.com/company/asknory/ Conor Sheridan on LinkedIn: https://www.linkedin.com/in/sheridanconor/