Francesca Roche sits down with Arian Zandi, Founder at Lumière AI, to examine how AI-native point-of-sale technology could help restaurants lift spend per guest while preserving the personal service guests value.
For many operators, the point of sale system remains a record of what has already happened, moving orders from the front of house to the kitchen and capturing transaction data for reporting, but the pressures facing restaurants now demand more immediate support.
Rising costs, constrained margins, and the limits of menu price increases have made every service interaction more commercially significant, with Zandi arguing that technology needs to help teams make decisions as those moments unfold.
“Margins are laser thin,” Zandi says, adding that technology embedded in day-to-day operations is becoming essential for restaurants that need to increase revenue in real time.
Much of the hospitality industry’s AI focus has centred on cost optimisation, reporting, and efficiency, yet frontline employees have a major influence on both the guest experience and the financial performance of a restaurant.
Zandi says: “The hospitality industry is not just reports, it’s not just insights and it’s not only managers that are taking action,” highlighting the role of waiters and senior chefs in shaping a service.
The opportunity is especially relevant as restaurants deal with high staff turnover, which can leave teams without the time or continuity needed to build extensive menu knowledge and confident service skills.
Lumière AI reports between seven and 16 percent AI-assisted sales growth in early point-of-sale deployments, alongside increased spend per guest across its pilot restaurants.
Zandi explains that the effect may extend beyond the recommendations surfaced by the platform, as staff become more confident in advising guests and making their own relevant suggestions.
“One thing that I have seen from our restaurants and multiple clients is that waiters start to even recommend things that are not necessarily part of the recommendation engine.”
The interview also explores the distinction between proactive and reactive recommendations, a critical consideration for customer experience leaders who want to avoid making a relaxed meal feel like a sales interaction.
Rather than prompting staff to lead with a suggestion before understanding what a guest wants, reactive recommendations respond to customer-led interactions, the stage of the meal, and the items being ordered.
“Every interaction that the customer initiates first with the waiter” can create an appropriate opportunity for a recommendation, Zandi says, helping the exchange feel “seamless” and complementary.
That design choice could matter as restaurants seek to use customer and operational data more effectively without drawing staff attention away from guests or making service feel artificial.
Zandi also points to menu engineering as a longer-term opportunity, using granular performance data to assess dishes, pricing, margins, menu placement, and presentation.
Watch the full interview to hear how AI-native POS tools could help hospitality teams build confidence, support more informed service, and uncover new revenue opportunities at the table.
