The supercomputer which is sharper than Watson – New Technology

Amelia is new to customer service. She answers in two seconds, solve customer problems in four minutes instead of the quarter, and our customers are completely satisfied. Amelia never go home.

     IP Softs Development Manager Ergun Ekici expect that machines can replace some human data, but they get harder to replace creative tasks. Photo: Jörgen Appelgren
     

For 15 years, the New York company’s IP Soft worked secretly with Amelia – a software based on artificial intelligence. The goal has been to provide a system capable of three things – to understand natural language, learning through natural language, and to use the learned knowledge to solve problems.

Just a few weeks ago presented the technology publicly, and last week was the management of IP Softphone on a visit to Sweden to meet with Swedish prospective customers.

Fokus located initially on the use of customer service and internal help center. During the past year, Amelia worked on samples of dozens of American companies. And already in January 2014 reported Gartner, who got to talk to the companies, if an increase in productivity that are not matched by similar technologies.

Gartner also notes that such technology can replace people. “In a company resolved 56 percent of customer problems without human intervention, and the time to solve the problem had decreased by 60 percent,” writes Gartner.

According IP Soft results were achieved after 30 days of training when Amelia listened to conversations between customer service staff and customers. Among examples of similar technology, Gartner takes up IBM’s computer Watson, who won over the people in the quiz Jeopardy in 2011, writes that the IP Softs cognitive engineering is a level above Watson.

Ergun Ekici, Development Manager at IP Soft, suggesting that Watson and Amelia would be a great couple, but would not confirm any concrete plans for cooperation. The combination seems to be interesting: Watson’s strength is to analyze very large amounts of unstructured data, and to find patterns and statistically probable answer. Amelia’s strength is to understand and learn through natural language.