Amazon Web Services on Thursday announced the general availability of Contact Lens, a set of machine learning-powered capabilities that transcribe and analyze customer calls to contact centers. Contact Lens is part of Amazon Connect, a fully-managed cloud contact center service.
Contact Lens promises to help call center supervisors understand customer sentiment, as well as trends and compliance issues that arise during customer calls. It lets users easily search calls based on the topic (such as calls from customers returning an item), customer sentiment (such as calls that ended with a negative customer sentiment score) or conversation characteristics (such as calls with long pauses, or customers and agents talking over one another).
After pulling up search results, supervisors can look through the transcripts and view a visual illustration of the call characteristics. Using natural language processing, Contact Lens can all uncover new issues, such as whether there are a number of calls about a mislabeled item on a store's website.
The service also uses machine learning to automatically detect and redact sensitive personally identifiable information, like names and addresses, from call recordings and transcripts.
Later this year, Contact Lens will be able to alert supervisors to issues arising while calls are in progress.
Firms using Contact Lens already include Intuit, Accolade, and DH Enterprise.
Contact centers are a clear enterprise use case for machine learning and natural language processing. Conversations with customers can offer clear insight into ways a company can improve its business, but the sheer quantity of calls and transcripts makes it hard to leverage that information.
At last year's AWS re:Invent conference, when Amazon debuted Contact Lens, AWS chief Andy Jassy said that Amazon Connect was one of the cloud business's fastest-growing services. Amazon Connect launched in 2017.
Other cloud players are also helping businesses use AI within their call centers -- last year, Google Cloud brought its Contact Center AI into GA, enabling businesses to deploy virtual agents for basic customer interactions.