The Customer

A major auto manufacturer manages an influx of customer issues and maintenance requests on a daily basis. The company’s top priority is ensuring a great customer experience for those dealing with issues. Immediately the customer inquiry must be routed to the appropriate group for assistance and manual routing is inefficient, error-prone, and expensive. Additionally an influx of customer inquiries often indicates a larger issue to be addressed by engineering, such as a faulty part.

To sort through these requests the company clusters them into various classes to then be passed on to the relevant party for follow up. This results in thousands of possible classes that change over time as new auto models are introduced and suppliers change. With so many classes, any error is not just a lost opportunity, but an additional cost. As mistakes receive push-back from dissatisfied service representatives assigned out-of-scope issues, customers issues go unaddressed resulting in higher costs. While the auto manufacturer’s team had already developed an AI model with a prestigious university to help with clustering, the model only reached 40% accuracy, igniting a search for tools that would produce models with greater accuracy.

The Solution

The key problem identified was the lack of communication on classes of interest.’s meta-learning allows it to incorporate a wider range of customer data. By asking customers to specify classes based on the data they submit, our platform learned what types of classes they would be interested in. Customer data is then analyzed using a variety of techniques to build a better clustering model. This approach led to a model accuracy of 90%, a 50% increase over the previous top-performing model.

The Benefits

The addition of an intelligent clustering model into their CRM system allows a level of efficiency never before possible. Customer issues are handled faster and more accurately than ever, and problems are correctly assigned to relevant processes the first time—leading to a much better customer experience for its customers who saw their issues resolved faster and with greater specificity. When it comes time to update the system, the platform nature of allows a new model to be made in moments without the need to import original data.

Create Your Free Account


Share this Post