IBM's Watson to Tackle Healthcare Next

NEW YORK (CNNMoney) -- IBM's Watson computer thrilled "Jeopardy" audiences in February by vanquishing two human champs in a three-day match. It's an impressive resume, and now Watson has landed a plum job.

IBM is partnering with WellPoint, a large health insurance plan provider with around 34 million subscribers, to bring Watson technology to the health care sector, the companies said Monday.

It will be the first commercial application of Watson, which is a computing system that aims to "understand" language as humans naturally speak it. IBM (IBM,Fortune 500) has been working on Watson for more than six years.

The goal is for Watson to help medical professionals diagnose and sort out treatment options for complicated health issues. Think of the system as an electronic Dr. House.

"Imagine having the ability to take in all the information around a patient's medical care -- symptoms, findings, patient interviews and diagnostic studies," Dr. Sam Nussbaum, WellPoint's (WLP, Fortune 500) chief medical officer, said in a prepared statement.

"Then, imagine using Watson analytic capabilities to consider all of the prior cases, the state-of-the-art clinical knowledge in the medical literature and clinical best practices to help a physician advance a diagnosis and guide a course of treatment," he added.

WellPoint plans to begin deploying Watson technology in small clinical pilot tests in early 2012.

Speed and natural language: Watson can sift through 200 million pages of data and provide a response in less than three seconds. But perhaps even more impressive than Watson's speed is its ability to process natural language, the way that humans speak it.

That's no easy feat for a computer. Human language is full of subtleties, irony and words with multiple meanings.

Take the "Jeopardy" example. Watson studies the questions by considering many factors, ranging from straightforward keyword matching to more complex challenges like homonyms (the bark of a tree is not the same as a dog's bark) and paraphrasing ("Big Blue" is the same thing as "IBM").

Watson is able to do this quickly thanks to software that runs on 10 refrigerator-sized racks of IBM Power7 systems. The machine is a grandkid to Deep Blue, the chess-playing IBM supercomputer that trounced world champion Garry Kasparov in 1997.

IBM said early on that health care is a field where it anticipated commercialization opportunities for Watson. Other markets IBM is eying include online self-service help desks, tourist information centers and customer hotlines.

Watson's "Jeopardy" face-off against champs Ken Jennings and Brad Rutter, which first aired February 14-16, will be re-broadcast starting on Monday.

Via: http://money.cnn.com/2011/09/12/technology/ibm_watson_health_care/index.htm?hpt=te_bn9

Coming soon: Call centers that don't suck

Just about every consumer who has dealt with a call center has faced the frustration of navigating through an unending sea of menu options or waiting eons to talk to a human being only to wind up with the least helpful employee.

Those annoyances and inefficiencies wear away at customers' satisfaction with the company they're dealing with, and that can ultimately cut into the company's bottom line.

But all is not lost. Real help may be on the way.

A new service developed by IBM (IBM, Fortune 500) and specialty insurance provider Assurant Solutions (AIZ, Fortune 500) plays matchmaker between customers and call center reps. IBM's service, called the Real-time Analytics Matching Platform (RAMP), connects customers with the call center employee best qualified to serve their individual needs.

Currently, most call centers only match an agent's availability and product mix expertise with the customer's particular product, making it an inexact science, to say the least.

"It's a little bit of a crap shoot regarding who you're going to talk to when you call a call center," said Toby Cook, analytics practice leader at IBM Global Business Services. "We combined our expertise in contact centers with our analytics to take it to another level. It's like speed dating for contact centers."

How it works
When a call comes in, RAMP instantly pulls up data about the customer, including past purchases and previous calls. The system then cross-analyzes that data with the skills of each call center agent.

Finally, RAMP uses its so-called "matching engine" to connect the customer with the best-suited agent within an acceptable amount of time.

For example, if a customer were to dial into a wireless company's call center, RAMP would be able to determine what phone the customer has, how much he or she uses it, what features the customer uses, and other identifying factors like the customer's age. The service would process all that information and match up the customer with a representative who has had past success with others who fit that criteria.

Traditional call centers might treat all customers who own the same BlackBerry models the same, but that can fall short when it comes to providing the best help. One BlackBerry user might be a global business traveler, while another user might be on a family plan and like to send a lot of text messages.

IBM's technology daily measures call center reps' success with a number of factors, including the number of customers they convince to stay with the company. RAMP also determines long-term success: some customers may be convinced to stay with the company after speaking to a rep, but then call back later to cancel.

All of RAMP's analytics are done in a matter of seconds. And since the system matches the analytics with data about customers' acceptable wait-time thresholds, IBM says the time customers spend on hold has either stayed the same or, in some cases, improved.
A success story

Though RAMP only became commercially available in the past few weeks, it has been through about six years of testing.

In 2004, Assurant Solutions took note of its call center issues and developed the analytics-based routing technology that underpins RAMP.

"We took data that has been there for a long time, but hasn't been exploited,' said Cameron Hurst, director of Assurant Solutions. "There haven't been a lot of new ideas in call centers in a long time, but we think this is a new angle on an old business."

Though IBM's version of the system is enhanced and portable, Assurant's more rudimentary system still had a drastic impact on the company's sales figures.

The help desk
Assurant determined that matching customers with the best-suited call center agents for their needs not only improved customer satisfaction, but it also boosted the company's sales. After the first year of using its new technology, Assurant's revenue from its call center rose 29%, and sales through customer retention grew 37%.

The system improved Assurant's call-center attrition rate too. As the system matched up call center reps with situations they would succeed in, agent morale improved. Like at many call centers, Assurant's call center reps are paid on commission.

"The more customers they save and the more they sell, the more money they make," said Hurst. "The system has a knack of delivering the right customer to the right rep, giving the reps the highest opportunity for success. They love it, and it has resonated very, very well with our customers."

For the original article, please click here.