IntelliResponse Systems, a provider of virtual agent technology solutions, has released IntelliSuggest, a new feature of IntelliResponse Virtual Agent 6.2 that leverages predictive matching technology to deliver answers to self-service questions.
IntelliSuggest's predictive matching capabilities match questions posed to possible answers in a company's knowledgebase based on the intent of the query, and not the keywords. By using this unique, proprietary approach, IntelliSuggest can find answer matches even when the typed keywords are neither in the title, nor in the text of the answer. Options that the user can select are based on match probability, with the order changing in real-time as users get more exact with their questions.
"A common challenge with existing search and Web self-service/virtual agent technologies is that their understanding of customer intent is not always clear or explicit, which limits the accuracy of responses," said Paul Smith, vice president of product and services at IntelliResponse, in a statement. "Many of these technologies rely on keyword matching, which drives the user to a set of results. The problem with this approach is that in most cases keyword matching only returns previous entries that were provided by other users based on exact matches of those keywords. Additionally, multiple responses are often ambiguous or confusing to the consumer compared to being able to deliver a single correct answer."
Empirical data gathered from live customer deployments using IntelliSuggest have far surpassed the company's expectations. A/B testing results have demonstrated up to a 17 percent improvement in answer accuracy and a more than 20 percent reduction in customer effort.