Intelligent CIO Middle East Issue 58 | Page 33

Q + A + Q + A + Q + A + Q + A + Q + A + Q + A + Q + A + DR NICOLA J. MILLARD, PRINCIPAL INNOVATION PARTNER, BT EDITOR’S QUESTION Machine Learning tends to work best in areas which are data rich, process driven, and unambiguous, for example cyberattacks on networks, diagnosis of very specific health conditions, pattern recognition for standard equipment, matching a defined set of products with a specific set of requirements. That said, no amount of secret sauce will revolutionise a business unless its data is in order. Machine Learning does not work by magic – it depends on data. If that data is unstable, inconsistent and spread among multiple legacy systems, Machine Learning becomes more difficult and far more expensive to do. Aside from data, enterprises also need to step back and ask what problem they want Machine Learning to solve and whether it will actually solve it. For example, if they want to deploy a chatbot to improve customer experience, does it actually improve it or just add another level of frustration for customers if the ‘bot’ hits a dead end and abandons them if it can’t understand what they want? Enterprises need to understand how Machine Learning will integrate with legacy systems and processes (e.g. in the contact centre), how much training it will require and what the business case is (given return on investment may take some time). There are some interesting opportunities around something I term the “me”conomy. This is all about leveraging customer data to tailor experiences. It starts with personalisation – something that Amazon and Netflix have done for years and customers like. The next step is proactivity – i.e. telling customers things that they might want to know (e.g. appointments, fraudulent activity, faults, contract renewals), on the right channel and at the right time before they have to tell you. This is generally regarded as a good thing by customers but could become irritating if done badly. Finally, you could then become predictive and anticipate what customers might need before they know that they need it. This can get creepy, though! If customers get irritated or creeped out, they are likely to stop sharing their data. No data means that Machine Learning is no longer effective. Going forward, Machine Learning will inevitably get more accurate as more and better quality data is available. As organisations move more into the digital space, 5G and IoT connect more things and better data standards are agreed, it can become more powerful. CIOs and IT leaders need to ensure that the wider business doesn’t get carried away by the hype around AI and Machine Learning. They need to work with the wider business to identify where deployment of Machine Learning is appropriate, likely to deliver ROI, and (above all) deliverable. There may be “ CIOS AND IT LEADERS NEED TO ENSURE THAT THE WIDER BUSINESS DOESN’T GET CARRIED AWAY BY THE HYPE AROUND AI AND MACHINE LEARNING. a need to completely re-engineer legacy infrastructure, data and processes, so CIOs and IT need to engage with the rest of the business to gauge their appetite to do this and ensure that investment budgets are realistic. www.intelligentcio.com INTELLIGENTCIO 33