The cloud is coming: an interview with Credit Suisse co-CIO Nigel Faulkner

Credit Suisse co-CIO Nigel Faulkner tells Computing that cloud is inevitable, even in financial services

Unlike many enterprises that manage to get by with one CIO, Credit Suisse Group has three co-CIOs, one of which is Nigel Faulkner, the CIO of the bank's investment banking division.

Faulkner joined the financial services giant in June 2010 after nearly nine years as CTO of equities at Goldman Sachs. After nearly two years as CIO of equities, trading and market data at Credit Suisse he moved to his current role in February 2012, where he is responsible for the support, sales, trading, research and primary banking systems that serve a global workforce of about 4,000. These systems are in Credit Suisse's main business hubs in New York and London, its business centres in Hong Kong and Singapore, and in numerous trading offices across the globe.

Ensuring good value from IT

Since taking on the job, Faulkner has changed the way his team looks at TCO.

"Most people think TCO refers to total cost of ownership, but that is not what it means - it means cost accountability. So [we ask] how do we make our technologists be accountable for the costs of technology for the firm in a way that is reactive to the business drivers at any one time, while also leveraging their technical expertise," he says.

Faulker claims that this new approach to TCO has led to massive savings since he became co-CIO.

"This [change in attitude] wasn't through top-down pressure but through bringing tools, advice, education and technical options to the group and making it clear that if you're an app owner, for example, your job isn't just about writing cool code and getting it into the application - your job is to also be accountable to the firm for the cost of that technology and also for things like operational risk that the technology presents to the firm, because really no one else can do that," he says.

In essence, Faulkner sees part of his role as making sure his team think long and hard about the value any proposed project might bring, and to insist in this he has implemented a scheme called "portfolio repositioning".

"It's to do with what is the right mix vertically versus horizontally implemented technology - and I think we've taken a slightly different approach to some of our competitors. We tend to err very specifically on the side of questioning whether something should be implemented horizontally and we spend a lot of time convincing ourselves that it would be a good idea before we do it," he explains.

Faulkner explains that the bank has several horizontal programmes such as its strategic risk programme (SRP) and the execution layer - the part that sits between Credit Suisse's clients and its exchanges that comprises things like client connectivity and market connectivity.

"We've rewritten some of our key components in there, including our order management system, which is called Agora, and through innovation in the code and with the infrastructure guys, we've had a couple of orders of magnitude of efficiency and reliability improvement - so that's been a good programme," he says.

Hiring a data scientist is hard, for us it's even harder

"The [risk] space and the big data spaces are the two in which I'd say there is a massive shortage of experienced senior technologists with financial services experience," says Faulkner.

By this, Faulkner means people who know how to apply big data graphs to areas such as risk.

"When you look at the skills data scientists have - some parts [of the role] are computer scientists, some part statistician and mathematician, some part economist and some part domain experience as well - so the idea that you can encapsulate that all in one person seems pretty unlikely when you look at the breadth of the field of our investment bank, let alone the broader remit of Credit Suisse," he says.

The firm looks at the big data space as a "team sport", says Faulkner, and it has recruited accordingly.

It is also working with an unnamed external partner on an online training course focused on big data.

So how can Credit Suisse use big data to its advantage?

"We've focused quite a bit in the last couple of years on the authorised conduct risk space, so that we can better predict and detect unauthorised trades, for example," he says. "The amount of data that is processed through the trading desk on any single day is daunting, so we've done a lot of work in that space including some fairly leading-edge stuff using natural language processing on electronic communications."

Faulkner believes instant messaging (IM) is a huge issue in the space because of the unstructured nature of the conversations.

"Companies have stripped the walls of what people use IM for but it doesn't mean it is as easy to digest as mail which is more like the spoken word," he says.

One of the ways the company wants to tackle big data challenges is by creating a "data lake".

"It will be a single source of our structured and unstructured data but in an ontology that our whole firm can understand. We only want to do that once and we want to do it on a demand driven basis so that then each business that wants to make use of big data analytics can sit on top of that data lake.... It provides a massive data asset for the rest of the investment bank to leverage," Faulkner explains.

A clearer path to the cloud

The firm uses what Faulkner calls an internal cloud offering - that is any combination of VM implementations. He calls it the "dynamic nature of allocation of resources".

But the company is now looking seriously at working with external vendors on pure compute.

"Today we have to provision to peak demands plus a percentage that we feel comfortable that we won't run out no matter what happens in the external world - it would be great if we could provision to a smaller percentage than that, so if we could figure out the intersection of the curbs between efficient use of all our compute grid and maybe that's 60 per cent of what we have today, and then externally for the rest, that would be a good optimisation of our datacentre space versus the cost of the compute grid versus the cost of bursting it out," he says.

This would still enable the company to do its critical calculations internally.

But the firm would still have to tackle the data privacy challenges it faces with the cloud.

"We're clearly not using cloud in almost any cases but I see it inevitable that we will, and there is enough desire and momentum within our industry and others, where we will be able to prove to all interested parties that it's just as secure to use external cloud as other forms of resources," he says.

Those parties include clients and regulators, and Faulkner claims that the firm has a huge duty of care over data of its clients - higher than any other type of firm. But he believes that the company can solve those problems.

"Cloud is coming, cloud is already around and we're starting to use it for anonymous compute, but I do see that we will use it for other things as well... but I wouldn't put a timeline on it," he state

The CIO role

Faulkner believes that the CIO role has evolved over the last four years as less of a pure service provider and much more of an integrated part of the senior business management team.

"When you look at how much of the strategy in the investment bank is predicated or handled and driven by the tech team it's the majority not the minority. The requirement is for the CIO to be very engaged with the management levels and active in helping them define their strategy on the back of that - that has grown exponentially over the last four or five years and that requires some more general management type skills than may have been traditionally the case for the CIO," he suggests.

And is there space for the chief digital officer and chief data officer, or is the role of the CIO slowly being broken down or replaced by other roles?

"I think they are all roles that are necessary. For me, [to say the CIO role could disappear] is like saying there won't be CEOs in the future - there will always be CEOs in the future and they'll structure their management teams underneath them in ways to help them to manage the business in the way they want to, and it's similar for the CIO.

"I can't imagine a future where there is not somebody who is taking a senior leadership role in a company that isn't generally responsible for technology that it implements - it doesn't make sense for that to happen, but that doesn't preclude other roles like the CDO for example," he concludes.