Big C's AWS-powered AI assistant serves 20M customers with natural language shopping

Thailand's retail behemoth Big C has unveiled an AI-powered shopping assistant integrated into its mobile app, built entirely on Amazon Web Services (AWS). The system uses multiple specialized AI agents orchestrated via Amazon Bedrock, allowing customers to interact in natural Thai language for product searches, meal planning, and discovering new items. Instead of relying on a single monolithic model, Big C's architecture chains together purpose-built agents—for example, one agent handles product lookup while another recommends recipes based on ingredients. This setup enables responses in seconds, personalized to the user's preferences and shopping history.

With 20 million active app users, Big C is betting big on conversational AI to drive engagement and sales in Thailand's competitive retail market. The assistant's ability to understand complex requests—like "find ingredients for a spicy basil stir-fry under 200 baht"—reduces friction compared to traditional keyword search. Beyond convenience, the multi-agent approach on AWS ensures scalability, low latency, and cost efficiency. This deployment signals a broader trend: regional retailers leapfrogging to agentic AI systems that blend search, recommendation, and planning into one seamless interface, potentially setting a new standard for e-commerce in Southeast Asia.