Using AWS tech, TensorIoT developed a system that monitors customer motion within a casino, from gaming tables to bars & restaurants. Using this new data, the casino can improve their rewards programs without encumbering customers with physical rewards card.
How can you use existing camera systems to recognize clients and deliver seamless rewards based on their play at a gaming establishment?
To better understand the behavior of the client’s customers, we leveraged our chip-counting system and Amazon Rekognition to run ML models on the client’s existing camera systems.
In the hospitality industry, knowing your customers well can make or break a business. Traditional methods of gathering information on customers often rely on rewards accounts. These accounts require customer buy-in, and often leave big gaps in the understanding of customer behavior and preferences.
The client, a leading gaming company, wanted to see how IoT and Computer Vision (CV) could help them learn more about their customers, and, in turn, reduce friction and improve the customer experience.
The client needed a smart system that could gather and organize data about customers, learning information that would help the client provide personalized perks during each aspect of the gaming experience. To accomplish this goal, TensorIoT needed to implement a facial recognition system that could detect metrics such as how long customers sit at tables or slots, how much they bet, and whether the customer is using any other casino services. Dim lighting in the casino, as well as the reflections from the games, provided unique hurdles in getting the facial recognition system to accurately detect faces.
TensorIoT built a machine learning (ML) application that worked with the client’s existing camera system. First, we tuned our facial recognition to work within the casino environment. Then, we added Amazon Rekognition, a robust, managed, out-of-the-box ML service. The solution was then combined with our Chip Counting application, which tracks customer betting behavior. This allowed our client to view all customer data in a single, consolidated application.