Astute Starts Speakin' Your Language

Astute Solutions has unveiled a new application for its RealDialog knowledge management system designed specifically to help contact center employees search for answers to customer questions. RealDialog Agent Assist relies on a linguistic engine that understands plain language and delivers consistent answers regardless of how the question is phrased. According to Lori Angalich, Astute's director of marketing, Agent Assist was developed because the company kept seeing that contact agents were having the same problems as the typical user of a Web site -- namely finding information quickly and efficiently. Agent Assist can be used by contact center agents or incorporated into a company's public Web site, and can be incorporated into any CRM platform. The application allows an agent to enter a customer's question verbatim and get a reply that cites a specific file or paragraph within a document, and guides the agent through questions to ask the customer. "The big differentiator is linguistic ability," says Liz Shaver, product manager for RealDialog. "It knows what it knows -- and knows what it doesn't know." Rather than relying on a keyword search that automatically returns every document that includes that word, Agent Assist allows questions to be entered in natural language, including colloquialisms and sentence fragments, and returns only the most relevant answers. Shaver uses the example of Water Pik, an Astute customer: the company uses "Water Pik" in every product it sells, which would gum up the works in a simple keyword search. "With other CRM applications, there's usually a keyword search, so a search for 'warranty information' might give you 100 documents you have to search through to find the answer [Agent Assist] can give you directly," Shaver says. Imagine, for instance, if a customer were to call in to ask if a defective product were still under warranty. A customer service representative using Agent Assist would simply type in that question verbatim, and then be prompted to procure specific information from the customer ("Get the serial number"). The system would then provide the agent with relevant, appropriate scripted directives for that caller ("Send the product in to be replaced," or "Here are six nearby locations that can repair it"). Agent Assist also stores customers' questions so they can be retrieved for later use, creating an ability to "learn" from past queries to generate better responses in the future. The solution can also have its knowledge base updated by contact center agents, without the assistance of the technology staffers. Answers on file can also be customized based on format: A stored answer for an email query might include Web links and document attachments; a stored answer for a phoned-in question would not. Another practical application of having a database of customer questions is that it collects "customer sentiment" information in customers' own words, without needing to do a survey. Shaver says that if, for instance, the contact center were getting a high volume of calls from customers asking how to pay their bills online, and the company didn't have an online bill-pay option, that feature could be flagged for marketing to evaluate. Shaver says that companies typically see a 50 percent reduction in training time when they use Agent Assist, since agents' time is spent learning only the software instead of having to master the ins and outs of the company's entire inventory of products.

Related articles: Feature: Speak Up! 11 strategies to ensure that you're hearing your customers loud and clear; natural-language dialogue allows for more open-ended responses. Feature: Avoiding the Speech Rec. Wreck Speech recognition has struggled to move into the mainstream, but these applications are picking up where touch-tone IVR leaves off. 12 Ways to Boost Productivity Natural-language search engines are enhancing contact center productivity. Feature: 10 Strategies for Customer Service Success Contact center managers can no longer define success by call duration alone; GlaxoSmithKline deployed Astute Solutions PowerCenter to more than 80 employees at 18 sites in its corporate and outsource contact centers. Feature: Online Self-Help Evolves Many believe that natural language technology is the key to the future. Siebel Continues to Fly Solo as a Leader The vendor leads all evaluated companies in Gartner's Magic Quadrant for CRM Customer Service Contact Centers, while more than a dozen players chart as visionaries; new additions to the report include Astute Solutions. Mobility's ''Killer App'' Speech technology will be the key driver of Web innovation -- primarily through natural language and its use in mobile search. Re:Tooling -- Tech Solution: Web Self-Service Business Problem: The call center is overwhelmed with basic, repetitive inquiries. Ask Jeeves Is Acquired by an E-Commerce Operator IAC is betting that Web self-service technology, specifically natural-language capabilities, will continue to mature.