CIOIN

Amit Gupta

VP-IT Services, Fidelity Business Services India

Executive summary

That Fidelity's perfectly functioning data management system wouldn't be able to meet the needs of the future was a concern. Thus came in a change, in the form of centralization of storage. It would make the organization more efficient by improving the amount of time it took to locate data by 60 percent and archived e-mail searches by 30 percent.

They say: if it ain't broke, don't fix it, right? Wrong. At least Fidelity Business Services India thinks so. Every 60 seconds, a Fidelity analyst somewhere in the world makes contact with a company to manage its funds. To support the organization's incredible need for data, IT created a data management system that met every challenge business threw its way. Fidelity could have decided not to mess with it, but that would have been the old Fidelity. "Typically, IT initiated a project when something failed. That's being reactive. So, we thought to ourselves: this is where we're today, we know the business projections, what do we need to do today to make IT work tomorrow?" Amit Gupta, VP-IT Services, Fidelity Business Services India, remembers asking his team. "When we reviewed our storage, we realized that while it worked today, it may not tomorrow," recalls Gupta. It was a good thought. And Gupta was prepared to do the shovel work required to re-structure an existing system.

Till early 2007, data generated by Fidelity India's business units was stored and managed using discrete technologies and traditional Wintel servers. As expensive as this approach to storage was, it was the right thing to do for the start-up, says Gupta. But not so much in the long run. The data its 7,000 staffers created meant Fidelity had scale to use technologies that would afford them a 40 percent cost reduction on every MB ofdata they created. "Can this be achieved for a 50-person company? The answer is no. In that case, we would have stayed with the traditional servers because for that scale that's the right solution," says Gupta.

The business' number one priority was the availability of data and its integrity. They had no direct interest in the cost but they were sensitive about it." As interested as Gupta was in these objectives, he had his own wants. He needed the project to reduce the cost of storage, ensure adequate capacity for the future and better the company's recovery capability. "We wanted to deliver quickly with fewer resources," he says.

He used the organization's reporting structure to his benefit. Once an obstacle, the multiple number of business users IT had to convince, became a crutch. Since they had to come to him, he whetted their appetite and created a groundswell of support. To do that, he met with individual business leaders, following his CEO's advice. Getting their buy-in before presenting the solution to a larger group also ensured that stakeholders were "very collaborative" and that they offered to make presentations themselves to their teams.

The planning, the insistence on business participation and the documentation all paid off. Today, Fidelity has a state-of-the-art data management system. Data centralization reduced the cost of storage. Not only that, it also improved data handling and security. De-duplication and compressions features have been used to optimize storage and efficiently utilize available space. The creation of fully redundant cluster of exchange servers with optimum load-balancing has led to a highly efficient infrastructure. Not surprisingly, Fidelity also met its green commitments thanks to the project. Another satisfied group in addition to business users, auditors and the finance department. This enterprise-wide satisfaction could be reason enough for other CIOs to tackle their existing systems. But be warned it's a lot of work persuading people to take on a devil they can't see. Just ask Gupta