It is no longer enough to capture static information on the enterprise, even if that information is kept current in frequently updated databases, according to Vivek Ranadive, chairman and CEO of Tibco Software. The corporation of the future will analyze information as it comes in to predict events.
"We will shift from static architectures to dynamic ones, based on the enterprise service bus," he predicted at the Tibco User Conference in San Francisco last week. It's not surprising he would say that, since Tibco offers an enterprise service bus. But even Tibco has a ways to go to meet Ranadive's definition of the future.
"What is the point of learning why we lost a customer three months after he's gone? Wouldn't it be better to anticipate events before they actually happen?" he asked.
A modern day, event-driven system would capture and "kick 100 million events" that occur in the enterprise infrastructure software in a day "up into the event cloud," where business analysis software would inspect them for what they revealed about customers" in close to real time.
"Wouldn't it be better to find out exactly the right thing to say to a customer as he steps off a plane, rather than after he's been waiting for bags for 40 minutes?" he asked.
While not exactly the real-time, event-driven panacea that Ranadive described in a keynote address, Tibco last week announced it was updating its enterprise service bus software to take it a step closer to the kind of system he described.
Tibco is adding Service Performance Manager to its ActiveMatrix set of service-oriented architecture products. Service Performance Manager monitors services to detect any latency in their response times, measuring it against a mean it has established. If one service is lagging, it can "proactively predict and take action" to solve a developing problem by bringing additional servers online to head off an interruption, said Matt Quinn, senior VP of engineering, in an address to the San Francisco user conference last week.
"Service Performance Manager focuses on a single problem -- managing the performance of services. It compares performance to the service level objectives you've been given" in service-level agreements, Quinn said. It can track service-level agreement violations and even take predetermined actions to "prevent IT problems automatically and behind the scenes," he said.