NTT Communications have announced that together with IPsoft, it will launch a highly automated cognitive agent service for natural conversation support from the summer of 2016.
The service will leverage artificial intelligence to smoothly interact with customers, in English and Japanese, and learn from experiences to successively upgrade the quality of its automated service.
The service will use cloud-based AI to automate initial responses to customers who contact a call center or enterprise/store contact point in search of support or to make a purchase. As an automated service available 24/7, it will enable user-companies to refocus employee workloads on higher value activities. The service will initiate actions to complete a caller request, for example, issuing invoices, sending emails or mailing documents.
The new service will understand and engage in human interaction by using cutting edge technologies including IPsoft’s artificial intelligence technology, Amelia, and advanced Japanese-language processing technology that NTT Media Intelligence Laboratories has been developing for some four decades.
The cognitive agent will enable customers to receive a range of standardized services at any time, 24/7, and without having to wait. In addition the automated interaction will also speed up the resolution of problems that need to be referred to a human agent by ensuring the full context of a conversation is shared when the caller is passed over to the live agent.
- Based on highly natural conversations with customers, the service will be able to complete a wide range of incoming requests as well as help to diagnose and resolve customer problems.
- In cases where a customer’s request is vague, the service will query them for more information to diagnose the problem fully. The service’s high level of accuracy will eliminate the need to repeatedly ask the caller for confirmations (for example, “If this is correct, please press 1.”). It also will generate questions flexibly, thereby freeing customers from being forced to follow a rigid rule-based dialogue.
- If a problem or request cannot be resolved automatically, the service will transfer the call to a live operator. Moreover, it will use AI to learn from the operator’s response and apply this knowledge when responding to similar situations in the future. This efficient learning cycle will enable the service to become increasingly efficient over time.
- The service will be able to initiate follow-up action at the caller’s request, such as revising information in an application, correcting a billing error or changing various settings by connecting to the user-company’s enterprise system.
NTT Com expects to improve the speed of resolving customer problems and requests by continually upgrading the system’s knowledge including dictionary databases for specific industries and purposes.
NTT Com also plans to adapt the service for other AI systems, such as the semantic search technology at its own customer support and contact centers, which is provided by Inbenta, a provider in natural-language processing and AI powered technologies.NTT Com also plans to collaborate in service development with partner enterprises, including companies providing consulting or CRM services.
Going forward, NTT Com expects to offer related services in various fields, leveraging both AI and the company’s highly safe and reliable cloud environment built with carrier-level quality.