Case Study

Smart Shopping Assistant

A leader Retail

One of the biggest supermarket chains in the Benelux with approximately 25,000 employees.

The Challenge

To make the shopping process more efficient and straightforward, the client wanted to develop a smart shopping assistant to help their customers better prepare for both online and offline shopping. They wanted to incorporate text and speech recognition in their online shopping experience so that customers could manage their shopping lists intuitively using Google Home and Google Assistant on mobile.

The Solution

ML6 were engaged to deliver an internal proof of concept (PoC) voice bot to test the practical useability of state-of-the-art technology in speech detection and cloud computing. This PoC had two components. First, a minimal Dialogflow shopping application took voice commands, interacted with the customer (taking context into account), and performed appropriate actions, enabling products to be added to a customer’s shopping list by text or voice via Google Assistant or Google Home. Second, in order to make the application intuitive to use and realistic in practice, a recommendation system predicted customer-specific brand preferences, based on their past shopping behaviour.

Results

This internal PoC demonstrated the technical feasibility of enabling Google Home to understand shopping list commands from voice recognition, and to respond with the correct actions. It also proved the successful integration of multiple building blocks, such as Google Home and Assistant, Dialogflow, Google Cloud Platform and the client’s data, opening up a whole realm of further opportunities to enhance the customer experience.

Partner