When J.P. Rangaswami, the global chief information officer of Dresdner Kleinwort Wasserstein, first started working for the firm as a consultant in 1997, the technology bubble was still inflating. He headed the efforts to tackle three pressing technological challenges at DrKW, the investment banking unit of Dresdner Bank: the introduction of the euro in 1999, preparation for Y2K and the implementation of the Bank for International Settlements’ core principles for effective banking supervision, issued in 1997. These projects, which touched on every aspect of the bank’s activities, turned out to be the perfect apprenticeship for a CIO. By the time they were finished in 1999, Rangaswami was “very aware of every piece of technology in every location in the bank,” he says.
That training didn’t prepare him for what was to come, however. Rangaswami joined DrKW as a director in its information technology department in 1998 and took over as CIO in May 2001, when the once-booming tech sector was in free fall. A few months later the September 11 attacks roiled financial markets, prompting massive cutbacks across the securities industry. As a result, he had to cope with a halving of his tech budget without sacrificing service.
For the 47-year-old Rangaswami, who began his career as a financial journalist editing his family’s weekly, Calcutta-based journal, Indian Finance, the pre- and postbubble lessons have been well learned. In today’s more straitened circumstances, he sees cooperation, not competition, as the smartest way to boost his firm’s technological capabilities. In 2001, for example, DrKW embraced open sourcing by introducing Openadaptor, an application integration software tool that the bank developed in-house. Openadaptor helps applications -- a trade capture system and a settlement system, for example -- communicate with each other. DrKW posted the software code on a Web site (www.openadaptor.org), making it available to other firms to use and adapt.
Rangaswami discussed his technology strategy in a recent interview with Institutional Investor Contributor Suzanne Lorge.
Institutional Investor: The technological and financial climates have changed a great deal since you first began working with DrKW.
Rangaswami: I have about half the staff I had in 2001. Currently, my permanent staff is about 1,000, and including consultants and suppliers, it hovers around the 1,300 mark. It was 2,000 when I took over. Our technology budget was about E700 million [$620 million], after amortizing capital expenditures, in 2001. In 2005 it is approximately E300 million.
How have you adapted your priorities to those diminished resources?
I’ve had to transform the skill bases required for the growth of the late ‘90s into those needed for management and sustainability during a sharp downturn. As accelerated reuse of code and a move to platform-independent operations created a focus on collaboration across regional and functional silos, the skills needed switched from star quality to teamwork.
What are your technology priorities?
There are three drivers. The first is to provide users with increased virtualization, separating the function of an application from a specific bit of hardware. Platform-independent storage and grid computing are two examples.
Today’s platform is hardware plus systems and applications software, with a high proportion of vendor lock-in. Tomorrow’s platform will not be vendor-driven but rather market-driven and will consist of ecosystems of virtual processor, disk and input-output system, supporting service-oriented software and powerful collaboration tools.
We have implemented storage-area networks, moving from platform-specific storage to a utility model that provides storage to multiple servers. You can raise or lower the storage associated with an individual or department without having to fiddle with the specific hardware. But it is still early days with true platform-independent storage.
What about grid computing?
Institutions like ours tend to have a number of applications that have very high computational requirements, especially for structured and complex derivatives trading. Each asset class used to have its own server farms to deliver the computational power required. We have begun to merge these farms into one large grid, allowing for greater adaptability and scalability while reducing overall costs.
And the second driver?
More and more of our applications are consumed as services, and we need to allow greater personalization for the individual user. A simple example would be what we did with our telephone directory and our organization charts and tables of authorities. First, we had a static impression of each, usually printed in hard copy. Then we moved them to a nascent intranet. Now we have all this combined into a service. You can use the application to find a person’s phone number, to e-mail or text them, to visit their blog if they have one and even to work out where they are located and where they fit into the organization’s structure.
And the third technology driver?
The final driver is the need for increased collaboration. From blogs or wikis [Web-based applications that allow users to view and edit content directly] to instant messaging, more sensible use of e-mail, video and much simpler chat tools, there is an increased demand and supply of these communications methods.
Blogging bankers?
We have more than 120 internal blogs and wikis within DrKW covering subjects ranging from information security and user-centered design through to credit markets and credit products. They represent conversations between individuals as well as groups, tossing ideas around and getting a larger population of skill to focus on a subject. The more the subject lends itself to diarylike entries, the more bloglike the outcome. Where greater structure and work flow is required, the more wikilike the outcome.
What new technologies are you experimenting with at DrKW?
There are three or four that merit mention. We are refreshing our technology infrastructure over the next 12 months to take advantage of the convergence between voice and data communications. An example would be that of the traditional trader turret and its interaction with hoot-and-holler systems, voice and data networks and market data applications. We are using advances in technology to drive more of the traditional turret functionality into software rather than hardware. We expect to gain in communications efficiency, in collaboration and even in disaster recovery as a result of this investment.
The second zone in which we will begin experimenting is radio frequency identification, or RFID.
RFID devices are attached to objects and transmit identification codes to a reader using radio waves. How do you apply that in an investment banking environment?
Initially, we intend to use RFID to track IT assets to manage our portfolio better, to track people to refine and improve environmental support and to alter device or user-access permissions more seamlessly.
A third zone we’re watching with interest -- we are experimenting, but we have not yet found resolution -- is contextual search, which uses the implied meaning of terms to refine output. For example, in banking, convertibles tend not to mean open-top vehicles. We’re looking for the right form of easily administered tools to help people search across the environment without worrying about where they filed something and what they called it. This is not trivial and leads to my fourth and final point.
This is the implications of collaboration and sharing for personal and confidential information. We need a model that treats customer information with the right level of confidence and does this in a secure way without creating a completely unusable set of search tools.
How does open-source technology fit into your strategy?
In 2001 we actually went for hosting an open-source site, www.openadaptor.org.
Open source is the new outsource. If I can off-load 50,000 jobs to people who don’t even work for me and then motivate them to provide value to this organization, this is a much smarter thing to do than to go through the major political complexities of offshoring and outsourcing arrangements. The commoditized work can be managed and tested by the open-source community, allowing my guys to focus on the business end of the value chain, such as how to use the emerging collaboration tools to generate business value. I think these are very exciting times.
