IBM's megacomputer Watson is not the only game in town when it comes to complex, deep learning technology for the enterprise.
"There are more cost-effective ways to enter the market and start applying those same disciplines to provide a better customer experience," said Dan Miller, founder and senior analyst at Opus Research.
Specifically, the use of virtual agents that harness artificial intelligence, machine learning and natural language to ascertain the purpose of a question and get the answers quickly. The difference is virtual agents tap into a finite data set rather than the unlimited sea of information available to Watson. While IBM's technology has its place in sophisticated applications in the health care and pharmaceutical industries, it can be overkill when it comes to customer self-service uses.
"Watson is great in a highly complex environment, but it was never designed to be low-labor intensive," said David Lloyd, CEO of virtual agent provider IntelliResponse. "If all you are trying to do is help a customer get to a finite set of the right answers, it is a much different process." The trick is figuring out the many ways a customer might ask the question in order to serve up the right information, he said.
Right now, more than 50 vendors offer virtual assistant technology, including Anboto, Expertmaker, Next IT, and Nuance. And the market could be about to explode. According to Opus, in 2013, enterprise virtual assistant spending lingered around $100 million. But as these virtual agents get better at learning new tasks and responding to customers in personalized ways, Opus predicts spending will approach $700 million by 2016.
STEPHANIE NEIL is a freelance writer and a correspondent for Business Information. Email her at firstname.lastname@example.org.