'85% of world's data is unstructured'

IN CONVERSATION

Stephen Gallagher, currently leading Accenture's business intelligence practice in Europe, has done extensive work in information management since 1985. In an interview with Barkha Shah, he explains how information management and predictive insight are becoming important enterprise tools.

What exactly is information management and why is it gaining momentum?

Analysis of structured and unstructured data is information management. We store a lot of data on computers today and these come in various forms - documents, e-mails, pictures, maps.

According to Gartner, by 2012, enterprises will need to handle 30 times more data than in 2002 and IBM states that an estimated 85 per cent of the world's enterprise data is stored in unstructured formats. Analysing unstructured data is not easy.

However, organisations are in need of such analysis because business conditions are changing, markets are maturing, and new regulations like Sarbanes-Oxley have come into existence. Business intelligence applications, therefore, are a priority for enterprises today.

How has the role of business intelligence changed over time?

Business intelligence was earlier analysis of historical data. So it could inform you, for instance, about the sales that happened last month. It looked backwards. Today, it involves real-time predictive data. It is embedded in the processes, tools and hardware that an enterprise uses and analyses real-time information.

How are information management services being used for predictive insight?

Accenture has developed a proof-of-concept for a super market chain. The trolley that a customer uses at a super market would have a window that would read the radio frequency identification (RFID) tag on each product that he would buy.

The application would therefore analyse the buying pattern and give a suggested shopping list to the customer. It could also enable a one-on-one marketing proposition for the customer, based on his individual needs, which would be more effective than a mass market one.

We have also completed a pilot programme for a transportation authority whose bus fleet carries more than one lakh passengers daily in Missouri and Illinois.

Sensor devices were placed on moving parts of the engines of the buses that sent information to computers at Accenture Technology Labs in Chicago. The captured data helped in predicting upcoming failures by analysing the engine and transmission information.

Where about India?

Our Bangalore team is at present working on a proof-of-concept for a global insurance company to analyse hidden trends with regard to underwriting. Predictive analysis can typically help an insurance company to predict frauds, claims pattern, or whether underwriting a certain risk is worthwhile.

We also plan to set up two centres of excellence for data migration and data analysis in India. Data migration takes 70 per cent of the time used in building solutions. This centre could be in Mumbai, though we have not finalised on it as yet. The centre for data analysis is likely to be in Bangalore.