Learn how SaladStop! partnered with 1CloudHub and AWS to deliver AI-powered personalized meal recommendations, achieving higher customer satisfaction and operational efficiency.
Benefits
Increase in customer satisfaction through personalized nutrition
Growth in repeat orders from enhanced engagement
Boost in operational efficiency via smarter inventory allocation
Higher order value driven by AI-powered upselling
Overview
SaladStop! is Asia’s first and largest healthy food chain, with over 80 outlets across Singapore, the Philippines, Hong Kong, Indonesia, Korea, and Thailand. Founded in 2009 with a mission to make healthy eating accessible and enjoyable, the brand offers more than 500 possible salad combinations and operates a multi-brand portfolio across the region. As customer expectations evolved toward personalized, nutrition-focused dining, SaladStop! needed to enhance its ordering experience and operational efficiency. AWS Partner 1CloudHub helped SaladStop! implement a Generative AI-powered solution on Amazon Web Services (AWS), enabling real-time personalised meal recommendations, improved customer engagement, and more efficient operations.
The Challenge
As customer expectations shifted towards highly personalised, nutrition-focused dining experiences, SaladStop!’s traditional menu and ordering systems were struggling to keep pace. Despite the breadth of their menu, the lack of intelligent personalisation was creating friction at every stage of the ordering journey.
Key Pain Points
🔹Manual bowl-building was a time-consuming and confusing experience for customers navigating dietary needs
🔹Customers lacked guidance on flavour compatibility, making personalised choices difficult
🔹Finding ingredients that matched cuisine preferences, vegan requirements, or allergen restrictions was not intuitive
🔹Digital ordering platforms offered limited personalisation, leading to increased drop-off rates
🔹Inventory planning was reactive rather than data-driven, limiting operational efficiency
To maintain its leadership in the Quick Service Restaurant (QSR) space and accelerate its vision of personalised nutrition, SaladStop! needed a scalable AI-driven solution — one that could deliver meaningful personalisation without adding operational complexity.
The Solution
1CloudHub, an AWS Partner, was engaged to design and deliver a Generative AI-powered ecosystem for SaladStop!, supported by the Infocomm Media Development Authority (IMDA) through its GenAI x Digital Leaders initiative. The solution centres on an intelligent chatbot — the LuLu Chatbot — that delivers real-time, personalised meal recommendations through natural language conversation.
80+
Outlets Worldwide
500+
Salad Combinations
14
AI-Orchestrated Tools
40+
FAQs in Knowledge Base – per brand
AI-Powered Chatbot and Smart Recommendation Engine on AWS
At the heart of the solution is LuLu — a conversational AI assistant built on Amazon Bedrock using Anthropic’s Claude model. LuLu engages customers in multi-turn, context-aware conversations to understand their dietary preferences, nutritional goals, and flavour affinities, then dynamically generates personalised bowl recommendations in real time.
How It Works: Multi-Layered Architecture
| Capability | How 1CloudHub Built It |
|---|---|
| Natural language ordering | Amazon Bedrock (Claude) for multi-turn intent recognition and conversational memory |
| Personalised bowl recommendations | Amazon SageMaker ML models using content-based filtering across nutrition, flavour profile, and allergens |
| Real-time ingredient availability | Live SQL queries against the ordering platform, filtered by location and timeslot |
| Semantic FAQ search | AWS Bedrock Knowledge Base + Amazon OpenSearch across 40+ FAQs per brand |
| Smart upselling | Contextual AI prompts driven by pricing logic and order history |
| Cart and session management | Persistent GenAI session state via AWS Lambda and API Gateway with WebSocket support |
| Data storage and retrieval | Amazon RDS (PostgreSQL) for structured data; Amazon S3 for order metadata and nutritional matrices |
| Performance optimisation | Amazon CloudFront CDN for edge caching of static content |
Tool Orchestration Engine
1CloudHub designed a 14-tool orchestration system that enables real-time, stateful conversations by synchronising user sessions, APIs, and backend systems.
Conversational AI Layer
Built on Amazon Bedrock using Claude (Anthropic), enabling multi-turn conversations with contextual understanding and memory across the ordering journey.
Recommendation Engine
Powered by Amazon SageMaker, delivering personalized meal recommendations based on nutrition, flavour profiles, allergens, dietary preferences, and historical order data.
Knowledge Base
Built using AWS Bedrock Knowledge Base and Amazon OpenSearch for semantic search across 40+ FAQs per brand, ensuring accurate and grounded responses.
Backend Integration
Built using AWS Lambda, API Gateway, and WebSockets for real-time interactions.
Amazon RDS (PostgreSQL) and Amazon S3 manage structured and unstructured data, including order data, pricing logic, and nutritional information.
Outcome
Since deploying the LuLu Chatbot, SaladStop! has seen measurable improvements across customer experience and business operations:
Higher customer satisfaction driven by faster, more engaging and personalised ordering experiences
Growth in repeat orders, with personalised recommendations increasing customer loyalty
Improved operational efficiency through smarter inventory planning and AI-driven automation
Uplift in average order value, as contextual AI upselling encourages premium ingredient selection
The solution has positioned SaladStop! as a pioneer in AI-driven personalised nutrition within the Asia Pacific QSR industry, and demonstrates 1CloudHub’s capability to design, build, and deliver production-grade GenAI solutions on AWS.
“Our customers want fast, engaging experiences when they order their healthy meals, and cloud technology has given us the power to deliver personalized and convenient ways to make eating well easier. By working with 1CloudHub and AWS, we’ve been able to set new standards in the food service industry—combining deep technical expertise with generative AI innovation to make personalized nutrition easy and accessible for everyone.” said Adrien Desbaillets, CEO, SaladStop!
AWS Services Used
Amazon Bedrock
Core GenAI layer — powers LuLu's conversational AI using Anthropic's Claude for intent recognition, multi-turn dialogue, and response generation
Amazon SageMaker
Hosts ML inference endpoints for personalised bowl recommendations, combining content-based filtering with historical order data
Amazon OpenSearch
Enables semantic search across the knowledge base, grounding AI responses in accurate FAQ content
AWS Lambda
Serverless compute for tool orchestration, backend logic, and real-time API responses
Amazon API Gateway
Manages RESTful and WebSocket APIs for real-time cart and order updates
Amazon RDS (PostgreSQL)
Stores structured order, user, and availability data with high-speed SQL query support
Amazon S3
Manages unstructured data including nutritional matrices, pricing logic, and order metadata
Amazon CloudFront
Edge caching for static content to optimise chatbot response times across regions
Next steps
If your organization is looking to deliver personalized customer experiences, streamline operations, or explore Generative AI on AWS, 1CloudHub can help. We design and implement AI-powered solutions, build intelligent recommendation engines, enable real-time data integration, and orchestrate scalable GenAI workflows to drive measurable business outcomes.
Contact 1CloudHub to explore how you can accelerate your journey toward AI-driven customer engagement and operational efficiency.
