The Great Recession, and why IT can't save the world

According to Mastek chairman and co-founder Sudhakar Ram, one of the key problems is aging, overly-complicated IT systems.

Ram's IT company specializes in deploying solutions for large and complex transformation programs such as the London Congestion Charging Scheme and Britain's National Health Service. He carries a personal passion for making our connected world sustainable, and has written extensively about the need to reinvent globalization, governance, education, consumption and how we measure success.

I sat down with him at ZDNet's New York headquarters to see what he thinks should be done to avoid the next financial catastrophe.

ZDNet: On the systems level, what happened during the global economic crisis?

The traceability of the original loan is not available in most electronic systems. Suppose I'm Lehman [Brothers] and I have a AAA loan on my books, and there's a problem in Stockton, California. With all of that slicing and dicing of loans, there's no way to trace that loan back. This is pure speculation and gambling.

Is it really a hedge at this level, or is it something else?

Human ingenuity is far, far better than systems' ability to catch up. There's a huge amount of overhaul required in the financial system, end-to-end. As the markets revive, and as companies want to play for the long run, they will need to overhaul their systems. There's too much legacy and duplication and systems doing the same thing.

"IT cannot save the world."How do you create more sophisticated systems that are far more agile and that reduce redundancy? Some of this is abandoned caution. If risk-management systems were able to track some of this, and present it to the board...this information is [usually] not even available to a board member. You can't hold them responsible, because they have no clue.

It's not negligence, it's a lack of information in some way.

IT cannot save the world. IT can only make it possible for someone to take the lead to save the world.

ZDNet: What needs to be done, then?

You should be able to trace back to an original loan. At every level of abstraction, you need systems that can trace back.

The risk-management systems need an overhaul. The risks associated with derivatives, all of them are based on a series of assumptions.

We need sophisticated risk-management. The moment I introduce a product to a company, I should have the system to manage that risk, or else not be allowed to introduce that product.

Almost 30 to 40 percent of critical decision-making systems are actually Excel spreadsheets, in large organizations. A spreadsheet is a very unhelpful way of describing critical decisions.

Nobody knows where the business logic is in a specific business process. Sometimes there is conflicting logic. Risk-management data requires granularity – not the kind required for financial statements. What people rely on finally is their spreadsheet. Whether it reconciles with another person's spreadsheet, they don't care. Everyone has their own small reality.

ZDNet: How can organizations approach these changes? Clearly, they can't just start over.

It cannot be a big bang approach because it will put a business at risk. Even at a mid-size bank, you can't just rip out and put stuff back in.

The thing that we've found [is that] the new SOA approach actually helps with a gradual transformation. The approach we have taken with legacy programs is to put in an SOA layer and a front end that can work with multiple back ends – a complete SOA-enabled platform.

"The incremental approach has been tried and it has failed."Typically any business we go to, we have about 20 to 30 major applications that are critical to the business, and about 100 minor applications. We try to create a roadmap over a three to five year period to make the cost migration and make a case for every system. We aim to create enough safeguards around legacy systems so that there is enough integrity around that data to keep it safe.

It's a faced approach, blind and systematic.

Most systems, nobody knows what is inside. These systems have been put in place and the people who knew what they were supposed to do are all dead or retired. That means there's enormous risk in ripping out these systems.

There's no right. There are fifteen versions of right. There's no right -- it's whatever is there.

How do you reconstruct? This is not an IT problem, this is a management problem.

In the U.S., we suspect that there will be $500 to $600 billion spent to modernize these systems over a seven year period. How do people take this on and cost-justify it?

An incremental approach has been tried over the last 20 years and it's failed. There's been more and more [of a] band-aid [approach]. It's pretty much a non-option for most companies. The companies that have ripped out systems and put in an SAP or Oracle have done well. But smaller companies have yet to handle the fundamental issues in their systems.

With a nuclear plant, it's got a 20 year shelf life. Nobody asks when you replace it after it degrades over time. If you don't, it will blow up in your face! But applications degrade much faster.

If business was fairly stable, applications would not degrade. But obviously, we're not in that business environment.

ZDNet: What's your roadmap for a CIO?

First, you have to understand the existing scenario. Then you have to create a targeted architecture. It's a high-level vision that you create for five years. Then you look at the application situation and create a roadmap – you're trying to trade impact with risk, find low-hanging fruit. You take those and migrate.

Establish ROI and managerial confidence. You may come up with things where the risk is too high – keeping it outweighs the risk of migration. Those you do three or four years in the program. Upfront you do the easy stuff, and reap the benefits for the organization.

There's no silver bullet, there's no simple answer -- it has to be crafted company-to-company. Wealth management, insurance, healthcare, public sector -- such as congestion pricing in London.

ZDNet: What's Mastek bring to the table?

We want to create an impact. Technology is merely an enabler.

What's important is what the business benefit is. We're not people that come in and tell them what to do -- we work in collaboration with customers. Once the vision is clear, then we work on architecting a solution.

We offer predictability in a process where there’s a high level of failure. It's by show and tell. We work very well with visionaries and early adopters.