IBM Watson will work in call center support services

In the United States alone, $ 112 billion is spent annually on the work of user support call centers. At the same time, half of the 270 billion hits remain unresolved. Almost everyone can remember many examples when a call to the support service turned into a long quest to listen to the recorded standard phrases and press the number buttons to communicate with a living person and find out that he himself doesn’t understand your problem.

The Watson supercomputer, which beats champions of quizzes and has already begun work as an oncologist-diagnostician in some hospitals in America, is quite capable of coping with this work, according to IBM. According to the company, almost two-thirds of the unresolved customer issues could be resolved if the call center staff could search for information faster. Such a search takes on average between six and nine minutes per call.

In the coming months , Watson will be tested in support services by the first IBM customers — ANZ Australian Bank, Royal Bank of Canada, consumer behavior research company Nielsen, Israeli mobile operator Celcom, consulting and analytical company IHS. Watson will communicate with customers of these companies through various channels - in chat rooms, by mail, through mobile applications. Some companies even intend to acquire voice recognition systems so that Watson can talk on the phone.

Watson, which can “swallow” tons of information about all the company's products and services in a matter of weeks or even days, from official documentation and specifications to reviews on forums and reviews on popular sites, and is also able to understand requests in natural language, can find the necessary information almost instantly.
')
IBM is already testing Watson in its own call centers, and the first results are impressive - the computer tells the customer the necessary information 40% faster. Unlike ordinary support service employees who are not highly qualified and often have to search for answers with keywords in the documentation or simply on the Internet, Watson takes into account the semantic links between data fragments, like a good specialist who grasps everything on the fly.

Joyce Philips, CEO of ANZ Bank, says that the idea to use IBM Watson came about a year ago. The bank intends to start using Watson's ability to consult on insurance issues. For a few seconds he will be able to study in detail all the client’s documents and find places in the insurance “armor” that are not sufficiently protected, or vice versa, with an excess. Instead of filling in a heap of long profiles, the client will simply send copies of documents to the bank and answer some questions in a chat or by phone in a free form.

Although work in call centers is not as respected as the work of a doctor, and not as spectacular as winning a quiz, it can bring substantial profits today. If the contract with the medical company WellPoint was primarily a matter of prestige and trust in the abilities of artificial intelligence, then in the support services Watson can soon become a “workhorse”.