In search of information for a story on business intelligence technology 2.0, I was informed politely by the head of the BI practice of a global IT provider that I was at least six years behind the times.
“Business intelligence 2.0 has been functional since 2003 or 2004,” said Kamlesh Mhashilkar, who heads Tata Consultancy Services’ Business Intelligence practice. Might he offer a short history of this field?
In the mid-1990s, business intelligence technology was in batch mode and segmented by department, according to Mhashilkar. The structured data was delivered for analysis at day’s end or month’s end. By 2000, as companies consolidated information from across their lines of business into one place, the business intelligence horizon expanded to enterprise-wide from departmental silos. By 2003 the push was on to deliver business intelligence, not at day’s end but as soon as possible — in an hour or the next 10 minutes.
“That is where BI 2.0 came into the picture: How can people get the information in near real-time, or right time?” Mhashilkar explained.
As this transformation to immediacy was going on, the amount of business intelligence information exploded to include not just what’s found in tables and data warehouses, but also the less structured text coming from the Internet and wireless devices.
Now comes business intelligence 3.0, which inevitably tries to add correlative data from more extraneous sources, plucking from voices in the marketplace, video streams from surveillance cameras, and the local and not-so-local news shows. All this ancillary information is mixed in with a company’s data stores in the blink of an eye. The sellers will tell you this kind of intelligence makes factories safer, customers happier and commodity traders richer.
The algorithms for making correlations between data have been around for a decade, and much of the hardware for much longer. But in the BI 3.0 world, the surveillance cameras that are standard equipment in retail stores, for example, will serve not only to nab shoplifters but also to recognize confusion on customers’ faces and send help.
“We are doing R&D on this,” Mhashilkar said. And it is not just retail stores where this business intelligence technology could bear fruit. Think of the improved customer service at amusement parks: The business intelligence technology would allow operators to track where a guest is going and trigger alerts for an express pass, perhaps, or an upgrade at the park hotel. “The cameras are already there. The only investment is from the software, which will analyze the images or video captured by the cameras, and just do a synthesis on that to allow much better decisions in real time,” he said.
Or not, because, as every shopper knows, lifting the veil of confusion assumes the salesperson can read your mind, and that sometimes is not the case. Most of the confusion on my face when I’m in a store reflects whether I really want to buy something I can’t afford. And the last thing I want is for some salesperson who’s been sent out by a computer from the backroom to clear that up for me. It’s actually an issue of privacy.
But, as Mhashilkar explained, “To be very frank with you, companies still haven’t crossed level BI 2.0. They are still struggling with the integration of the data. They are still struggling with the correlation of the data in batch mode, and still trying to get near real-time intelligence.”
That’s good — at least for me, because I want the right to remain confused.