Toshiba Digital Solutions, which is using Artificial Intelligence (AI) to transform workstyles and make life more comfortable, is implementing operation process reforms that leverage communication in a variety of fields. One of these is "RECAIUS Contact Center Plus", a solution that supports contact center operations, which in recent years have been growing in importance as a bridge connecting companies and customers. In this solution, Toshiba brings together the speech technologies and natural language processing technologies which have been cultivated over many years, and enhances the AI technology that FAQs can be effectively utilized as knowledge, helping to improve the operator's response quality and customer satisfaction. We are taking on the challenge of going a step beyond business efficiency improvement to fundamentally transforming contact center operations. In this issue, we introduce our initiatives that provide support for contact center operations.
Contact centers directly connect companies and customers. They provide customers with the information they want, when they want it. Support response that generates a high levels of customer satisfaction turn customers into fans of companies and services. Customer feedback is extracted from the support response and used to discover seeds of future business. Products and services are becoming more diverse while consumer needs grow more complex. Contact centers, and the functions they provide, have become vital parts of corporate activity.
However, due to Japan's shrinking working-age population and the labor shortages, contact centers, like other businesses, are finding it difficult to secure sufficient numbers of operators. The greater diversity and advancement of products and services also makes operation more complex, creating a need for further operator training and career support. In order to develop contact centers as creators of strong bonds between companies and customers, measures need to be introduced that address these issues while creating profound satisfaction for both operators and customers alike.
Eyes are turning to technological advances and the use of AI. Business processes can be improved by entrusting machines with matters that they can take care of. The goal is to create environments in which operators can focus on providing hospitable support response, and quality can be raised while only using a small number of people. There are hopes that personnel can be provided with effective support in various aspects of contact center operations through the use of AI. This can be accomplished, for example, by combining machine learning with elemental technologies such as speech recognition and natural language processing.
However, there has been a high hurdle to applying AI to the communication field, an integral aspect of contact centers, unlike regular processing work in back offices and similar areas. There are many synonyms and homonyms in Japanese language, as well as a wide diversity of ways of expressing emotions. In order for AI to be used for contact center operations, it needs to be able to accurately understand the meaning of what is being said by the person on the other end of the telephone line.
Toshiba Digital Solutions has been at the forefront of improving contact center operation processes from the perspective of communication rationalization using Toshiba communication AI RECAIUS. One of the fruits of its efforts is the "RECAIUS Dialog Agent", which fuses our most-advanced speech recognition technology with speech synthesis, dialog, intent comprehension, and other technologies. It improves the efficiency of telephone contact point operations by automatically handling telephone calls.
This interactive voice response service accurately understands the meaning of what customers say over the telephone and uses synthesized voice to reply. Unlike conventional speech guidance, which uses the push-buttons on telephones, this service can smoothly confirm caller intentions and handle various procedures, alleviating the burden placed on customers. The service can place calls and engage in dialog in place of operators, supporting telephone contact point operations in a wide range of fields.
We are also making progress with an innovative solution that goes to the one step beyond improving the efficiency of specific operations such as telephone contact point operations to use AI to provide total support for contact center operations. That solution is “RECAIUS Contact Center Plus”.
As AI is further introduced in contact center operations, it is expected to help visualize the contents responded via the telephone by operators and to automate and standardize a wide range of operations. RECAIUS Contact Center Plus uses speech recognition technology to convert telephone conversations into text in real time, comprehensively visualizing what is said by customers and operators. Furthermore, it also has a function that learns about the relationships between the conversations and the FAQ entries referred by operators, and provides automatic FAQ entry recommendations. This improves operation efficiency and support quality and helps create greater customer satisfaction. It meets the various operation improvement needs of contact centers and converts conversations between operators and customers into value.
Let's look at the technology involved in RECAIUS Contact Center Plus's use of Frequently Asked Questions (FAQs).
In contact center operations, FAQs hold the key to greater customer satisfaction. Operators must use FAQs, which consist of organized and categorized frequently asked questions, to accurately respond to customer inquiries. Veteran operators use their experience to smoothly search for and use appropriate FAQ entries, but for new and inexperienced operators it is extremely difficult to search for FAQ entries while dealing with customers.
This is why we have developed highly accurate speech recognition technologies, technologies for using AI to recommend FAQ entries based on the contents of conversations, and technologies for extracting AI learning data from operator utterances and operation histories and automatically updating models. These technologies make it possible for FAQ recommendation accuracy to be improved automatically the more operators and supervisors (who monitor and support operators) use RECAIUS Contact Center Plus, without leaving this task to engineers of AI or analysis. If recommendation accuracy improves, even new operators will be able to appropriately and rapidly leverage FAQs, just like veteran operators (Fig. 1).
Until now, creating and updating appropriately categorized and useful FAQs required advanced analytical work by experts. This process was time consuming and costly. Because of this, FAQs were often difficult to update, and became obsolete or inaccurate over time. We often heard comments from the worksite expressing the desire to continuously add to and update FAQs in real time so that the latest inquiries could be handled.
That's why we developed technologies for understanding the contents of inquiries using speech recognition and natural language processing technologies, and automatically grouping (categorizing) those inquiries based on their degree of similarity. These technologies are applied to "RECAIUS Knowledge Editor."
RECAIUS Knowledge Editor extracts potential new FAQ questions from large inquiry histories, identifies important phrases and key words, and displays them as related information. This makes it easy even for people without analysis expertise to create and manage FAQs.
RECAIUS Contact Center Plus seamlessly integrates with RECAIUS Knowledge Editor and makes it possible to prepare high quality FAQs that reflect the latest customer feedback.
Toshiba Digital Solutions will use various AI technologies, focusing on RECAIUS Contact Center Plus, to support the operations of contact centers, which serve as bridges between corporations and customers, so that operators can provide high-quality service and develop greater customer trust regardless of their level of experience.
We will use the knowledge and experience we gain through these efforts to expand the use of AI in communications in a diverse range of businesses and operations (Fig. 2).
* The corporate names, organization names, job titles and other names and titles appearing in this article are those as of July 2019.