Winning with AI
Magalu Pioneers AI-Powered Shopping on WhatsApp
Magalu Pioneers AI-Powered Shopping on WhatsApp
The retailer is redefining the shopping journey—meeting customers where they are with simple, smart conversations.
Overview
Long before generative AI entered the mainstream, Magalu was already on a mission to become the Brazilian retail leader in the technology. The company was an early pioneer in using AI to improve customer experience and operational efficiency, from inventory management to marketing to sales.
Now, Magalu has raised the bar again. With the launch of the world's first end-to-end AI commerce experience inside WhatsApp, customers can search, discover, compare, buy, and receive post-sale support—all through a natural conversation with Lu, Magalu’s iconic virtual persona. “This is the dream of decades,” says CEO Fred Trajano. “The idea of customizing a store window for every consumer has been discussed for years. Now, it’s finally becoming real.”
Outcomes
3x
higher conversion than the app
4.5M
unique visitors in the first six months
85-pt
Net Promoter ScoreSM
The AI-powered shopping assistant, Lu’s WhatsApp, was developed through a collaboration between Luiza Labs (Magalu’s technology division), Meta (WhatsApp’s parent company), Google, and Bain & Company. Early concept testing with consumers from diverse backgrounds revealed a clear pattern: Shoppers wanted to find products faster, surface more affordable options, and replace complex navigation flows with intuitive conversations.
At its core, Lu’s WhatsApp runs on a large-scale language architecture that integrates proprietary and external models into a coordinated multi-agent system. “We don’t want to just use models built by others,” Trajano explains. “We train our own models based on our data. That’s our competitive advantage.”
Trained on data from millions of active users, 300,000 sellers, and more than 37 million published listings, the system operates through more than 20 orchestrated AI agents. These agents coordinate product discovery, recommendation logic, search interpretation, customer support, and transaction flow—delivering speed, contextual relevance, and consistency at scale.
Conversations with the company’s virtual influencer, Lu (inspired by Magalu founder Luiza Trajano Donato), fuel the entire shopping journey, from product search to customer support. Lu understands what customers want, interprets their intent and context, and responds in a way that feels engaging. Customers can use text, voice, or images to search for products, receive personalized recommendations, complete purchases and pay directly within WhatsApp, and track orders.
For example, a customer searching for floor lamps may be prompted with follow-up questions about style, budget, and intended use before receiving a curated set of relevant options—replicating the experience of an in-store advisor within a digital conversation.
Operating at this scale requires disciplined, data-driven management. The platform was designed for rapid iteration, scalable performance, and cost efficiency. Every interaction is monitored through performance metrics and quality controls to ensure reliability and continuous improvement. As Magalu scales the platform toward 33 million users, its objective remains clear: Combine marketplace scale with AI-driven personalization to redefine how digital commerce works—making it conversational, contextual, and frictionless.
We wanted a 100% conversational, 100% AI commerce, 100% automated experience — and that’s exactly what we created.
