March 17, 2008 09:29 ET

MADE IN IBM LABS: IBM Software Finds Hidden Product and Service Insight in Customer Interactions

NEW DELHI, INDIA and BANGALORE, INDIA--(Marketwire - March 17, 2008) - Researchers at IBM's (NYSE: IBM) India Research Laboratory have developed advanced software technology that uses sophisticated math algorithms to extract and deliver business insights hidden within the avalanche of information gathered by companies during customer service calls and other interactions.

The new business intelligence technology, called ProAct, helps organizations gain new business insights that can be used to improve customer satisfaction, develop new products and services, and find new business opportunities, which is the aim of IBM's cross-company Information on Demand strategy.

Business intelligence plays a critical role for companies seeking to make timely decisions, improve performance, and increase customer satisfaction to stay competitive. For most companies, valuable information that can help improve their businesses is often buried within the e-mail, text messages and transcribed call logs used to document interactions between customers and service agents.

ProAct, developed by IBM India Research Laboratory, is a text analytics tool, which automates previously manual analysis and evaluation of customer service calls and provides timely insight to help companies rapidly assess and improve their performance to maximize customer satisfaction. Designed to help dramatically improve call centers, this unique IBM technology mines both structured and unstructured data, giving unprecedented access to information to change the way call center agents work.

"Call centers are really giant data factories of market research: Every time a phone call, e-mail or text message comes in from a customer, it is automatically stored and in many cases either left sitting on a hard drive or being poured over manually," said Dr. Daniel Dias, director, IBM India Research Lab. "Most companies aren't even scratching the surface for what could be a gold mine of information about their products and services."

ProAct goes beyond the limitations of conventional customer service analysis tools to provide an integrated, objective analysis of structured information such as agent and product databases and unstructured data such as email, call logs, call transcription to identify reason for dissatisfaction, agent performance issues and typical product problems. This approach helps reduce subjectivity and guesswork compared to a manual analysis technique.

ProAct also can help automate a call center's tasks, enhance call center agent performance and identify new or expanded sales opportunities.

"Effective use of advanced text analytics, machine learning and natural language processing techniques enabled ProAct to bring forward useful business insights from customer-agent interaction data which helps increase process efficiency and better meet customer needs," said Dr. Shantanu Godbole, lead researcher on the project. "ProAct successfully reduced the customer-agent interaction data analysis time from 10 minutes per enquiry to only 30 seconds for the entire record of data."

Based on the Unstructured Information Management Analysis (UIMA) framework that IBM contributed to the open source Apache Software Foundation in 2006, the ProAct technology was initially developed as a service engagement. Now the new algorithms are being packaged in software and deployed in many IBM call center customers around the world.

UIMA is an open source software framework that helps organizations build new analysis technologies that help organizations gain more insight from their unstructured information by discovering relationships, identifying patterns, and predicting outcomes. IBM uses UIMA to enable text analysis, extraction and concept search capabilities in other parts of its portfolio of enterprise search software products, including OmniFind Enterprise Edition, OmniFind Analytics Edition, and OmniFind Yahoo! Edition.

ProAct builds upon other successful technologies recently developed by IBM Research to help companies use information as a competitive business asset. Through these Information on Demand technologies, organizations gain new ways to use text analytics, semantics, enterprise search, and voice recognition for analyzing all kinds of content, including text, voice and even images and video.

About IBM India Research Laboratory

The IBM India Research Laboratory (IRL) was established in April 1998 in New Delhi as the eighth of IBM's research labs. The lab expanded to its second site, Bangalore, in August 2005. IRL has a world-class team of researchers who have been recognized for their thought leadership and have developed innovations that help differentiate IBM both in the region and globally. For more information, please visit

Contact Information