The Old World Disrupted


I had the pleasure to speak at this year’s ABSL conference, which is quickly becoming one of the most important meeting points of business and technology in Central Europe. Representatives of cutting edge companies and important figures in business and politics (including David Cameron as the main speaker) met in Łódź to discuss various disruptions caused by modern technologies. I had the honor to represent GFT, and talked about the rise of Artificial Intelligence and its implications for IT service providers and the financial sector.

Recently, there have been many discussions on how Artificial Intelligence (AI) and a broader range of innovative technologies are disrupting the financial industry and IT service providers that support it. I wanted to shed some light on two major disruptions we are dealing with at GFT Poland, but such that are not actually caused by AI. We rather hope that AI will be able to help us stay dry while getting through them. I also wanted to describe how we handle these disruptions with our global Applied Technologies innovation network, and why I think it is a harbinger of positive changes happening in Central Europe’s IT services industry.

Marcin Kowalski, GFT Poland

Two Disruptions

The first disruption comes from our clients – financial institutions. These companies are under heavy pressure from regulatory bodies. Although regulation is not the first area that comes to mind when considering innovative technologies (AI included), this mostly applies to a category of regulations designed to prevent banks from abusing the financial system. Solutions for preventing banks’ customers from committing crimes are way more involved. These include KYC (Know Your Customer) due diligence and AML (Anti Money Laundering) prevention systems designed to keep terrorists and other criminals off the banking system. Such solutions are inherently very complex and process vast amounts of data with subtle correlations. They are incredibly demanding to manage by means of manual human work, which in turn generates massive labor costs. Even heavily overcrowded AML-analyst desks are not performing up to expectations. One of the reasons behind this inefficiency is that people are afraid of letting terrorists through. This leads to massive over reporting and generates even more work for other humans. This is why we have high hopes concerning potential involvement of AI in AML processing. Obviously, our clients’ interest is a trigger for GFT’s interest in AI. For example, we are currently working on extending AML workflows by means of AI in order to boost productivity, as there are numerous areas where AI outperforms humans.

The second disruption lies in the macroeconomy of the Central European region. Wages in Poland have been increasing steadily for years. This is not as extreme as in India, where salaries used to surge 15% a year on average, but the competitive advantage of Polish delivery centers based on labor cost is, in fact, melting. There is still a way to go, but clients are starting to ask what the advantage of Poland is, compared to more cost-effective geographical regions. We have time zone alignments and cultural compatibilities, but the world is becoming more globally aligned anyway. As GFT has several software delivery centres around the world, we need to keep an eye on what the competitive advantages of each one of these are. For Poland, there is a growing pressure on how we are going to climb up the value ladder to justify our cost.

How we responded

Hence the situation is as follows: our clients need and want AI, and at the same time, we can use it to improve on our value proposition.

In the case of core business – creating markets and winning contracts – delivery centres are out of the equation. One can hardly do business with Wall Street from Łódź or Poznań. Innovation however, and that includes AI, is something that we are able to deliver regardless of physical location or position on the corporate strategy ladder. Even though it does not seem to be always considered a “serious” business, innovation introduces a very different kind of efficiency. It is about coming up with ways to make business itself more efficient, instead of just providing cost effective labor. According to Gartner’s “An introduction to pace layering”, 75% of requirements for innovative systems come from business and only 25% are technical. These findings suggest that innovation not only enables service centers to work on more exciting assignments, but that it also encourages IT consultants to venture out of their familiar IT territory, to work with actual business more closely. This strongly applies to AI, which, without proper business applications, is just a headless horseman. Moreover, innovative assignments are usually run in a time and material manner and utilise open source tools, something very much aligned with habits of many software delivery centers.

This is one of the motivations behind the creation of the Applied Technologies framework at GFT. The idea is simple: to plant innovation hubs at every GFT location and enable employees to participate in development of machine learning, AI, blockchain and other innovative technologies. The key feature here is that it is applied equally to each of GFT’s locations. This creates a great opportunity for delivery centers like GFT Poland which excel on integration technologies, small and big data, testing or UX, but to a smaller extent on innovation. Contributing to this new type of efficiency has already enabled GFT Poland to improve its image internally towards its employees and within the entire Group as well. We now have consultants who built their global recognition by involving themselves in offerings, participating in client pitches and, generally, working on tasks that force them to jump out of the box of their usual IT routine.

One might be skeptical about IT service centers being able to deliver artificial intelligence solutions, but a closer look at the IT landscape in Poland reveals that the vast majority of our specialists are masters of science, with PhDs here and there. These people have strong academic background required to grasp complicated concepts associated with artificial intelligence. We are not risking overutilising them by making them work on AI – we rather risk them being underutilised, unless we enable them to do things that are more challenging intellectually. The potential of Central Europe has already been noticed by organisations such as ceai.io, goodai.com and others. This should not come as a surprise, provided that the cost of employing an AI specialist in Poland is way lower than of their counterpart in the Silicon Valley. In this area, the increasing wages of Polish IT specialists cease to be any kind of an issue.

Wrapping up

The fourth industrial revolution is already happening – and like all revolutions that preceded it, it will end with winners and losers. The question is which side we are going to find ourselves on, and I believe that organisations have the means to choose their fortunes. Leveraging this to a regional level may seem overly optimistic, but I strongly believe that enabling our employees to use as much of their actual potential as possible is both beneficial and inevitable in the long run. Applied Technologies enabled it globally and is a stepping stone towards high value services becoming a commodity in the Central European region. We are more than capable of being up to the challenge.

Post a Comment

* indicates required

Comment Area

  1. Eli28/06/2017

    Does AI help lowering operational costs if we factor in costs associated with building and maintaining AI platform? Another concern with AI automation is that of security hacking. Once AI framework is comprimized, it will need extensive and costly
    remediation.

    1. Marcin Kowalski05/07/2017

      Basically, same rules apply as for any other IT system. AI is not more expensive to develop nor more prone to hacking. AI seem to be actually less hackable as it is kind of tricky to build a model for a evil purpose as one needs appropriate data sets to teach a model for a specific purpose.

      I agree though, that “intelligence” in AI does not unfortunately mean it is in any way more resilient compared to other software.