OXXO
Computer Vision and AWS to test product suggestions inside a physical store during Store of the Future.
Data & AI · Amazon Web Services · Retail
The challenge
Every year, Oxxo organizes the "Store of the Future" event, where they present advances in retail store technology. As part of this project, iNBest.cloud was invited to collaborate and the development team set out to create a system that could be applied to the store of the future.
With the goal of improving the shopping experience, iNBest.cloud applied Artificial Intelligence (AI) tools and Computer Vision. The idea was to identify in real time the sex, age, mood, and facial characteristics of customers, such as glasses, beard, and mustache. With this information, personalized product suggestions could be made for each customer.
The solution
The iNBest.cloud team developed a solution that would change the way payments are made in physical stores. Through the use of Computer Vision and AI tools like Amazon Rekognition, customer data could be identified and analyzed quickly and accurately. This data could be used by the cashier to offer product suggestions that align with the detected characteristics, which in turn could increase the purchase ticket.
The solution's functionality allows this data to be private and only accessible to the cashier, if the company desires, respecting customer privacy.
Suggested products are fully customizable and can be configured according to customer preferences. The tool can be fed with a customer-authorized catalog, including their own brand, agreements with other companies, exclusive products behind the counter, or even slow-moving items. Artificial intelligence suggests products based on analysis performed through Computer Vision.
iNBest.cloud's solution is supported by Amazon Web Services, such as Amazon Lex, Amazon Lambda, AWS DynamoDB, and AWS API Gateway, which guarantees speed and the ability to display data in real-time, in microseconds, for both the user and the customer.
Results
- Shopping flow with more relevant suggestions
- Better cashier and customer interaction
- More sales opportunities through contextual suggestions
- Data displayed in real-time in microseconds