Meet Lulu, SaladStop!’s AI ordering assistant
It will debut in Singapore in November before its global launch.
Salad Stop Pte. Ltd. (SaladStop!) is unveiling Lulu, a generative artificial intelligence (genAI) assistant to help customers personalise and order meals through chat.
Launching in November, Lulu will be integrated into the app of the Singapore-based fast-casual restaurant chain as a chatbot that can process meal preferences based on calorie targets, flavour profiles, and dietary needs.
The system uses eight years of customer order history and ingredient combinations from SaladStop!’s 80 restaurants worldwide to recommend custom bowls in seconds.
“Customers need to tell us which restaurants they are picking up from and the time,” SaladStop! CEO Adrien Desbaillets told QSR Media. “It checks if the restaurant is open and the available ingredients. From that point on, we’re able to build the bowl.”
Lulu lets customers refine their orders through chat, offering nutritional guidance and suggesting substitutions before sending the final order to SaladStop!’s pre-order system. Loyalty points and payment options are integrated into the process for a seamless checkout.
Desbaillets said the system is designed to lighten staff workload by handling detailed ingredient and nutrition queries, freeing employees to focus on speed and hospitality. For customers, Lulu simplifies the complex “build-your-own” process and supports voice-based ordering for hands-free convenience.
It can generate balanced recommendations for complex orders—say you want high protein but low calories—without users needing to browse or calculate manually, he said.
However, Lulu won’t replace professional advice from a doctor, nutritionist, or trainer, Desbaillets said.
“So if you tell Lulu that you’re feeling really tired today or your body’s aching, or you upload your blood test results, we’re not here to translate that for you,” the CEO said via Zoom. “That’s not our role. If you give us specific details from a nutritional standpoint or flavour profile, then yes, we can suggest something.”
Future updates could let Lulu build meal plans developed with medical professionals, nutritionists, and fitness trainers, he added.
SaladStop! chefs trained the AI to understand flavour combinations, avoiding mismatched pairings that have plagued other quick-service restaurant experiments with generative AI in the US.
The challenge is tuning it right—too much data can make it repetitive; too little and it becomes random, Desbaillets said.
Lulu builds on SaladStop!’s existing data infrastructure, which includes years of order analytics and a robust ingredient database.
The rollout is a natural evolution of customer experience, predicting that conversational AI interfaces will become standard in the food industry.
“The idea is really how we extend this into new revenue streams and new customer interactions,” Desbaillets said, adding that he expects Lulu to boost revenue by 10% to 20%.
The AI assistant will debut in Singapore, where it will be refined until December before being launched across SaladStop!’s international markets, including the Philippines, Hong Kong, Indonesia, South Korea, and Thailand.
Desbaillets said local market testing would be key to customising Lulu’s recommendations, with each region having different health priorities. In some markets, customers focus on calorie counting; in others, it’s about protein or balanced nutrition, and he expects Lulu to evolve with that.
