By Meg Ramsey, Senior Director of Innovation Programs, Sungard AS
Machine learning is clearly good for business. The Accenture Institute for High Performance found that at least 40 percent of the companies in a recent survey had already implemented machine learning to improve sales and marketing performance. Those using the technology reported that new leads, upsells, and sales cycle times improved by a factor of two, at a minimum.
Yet, implementing machine learning requires people with the right skills. While many people have a general knowledge of machine learning tools, there are fewer that can build machine learning systems that work at scale. Such people must understand the mathematical and statistical basis of machine learning, have the programming skills to execute their ideas, and be able to write code that can be used in large scale production systems. These skills are hard to find.
A survey of over 200 IT decision makers conducted by Sungard Availability Services (Sungard AS) highlighted “Machine Learning Engineer” as the IT role seen as being the most difficult to hire in 2017, with 21 percent of the IT decision makers picking this as one of their top three concerns. In a world where autonomous cars seem poised to hit the streets, organisations don’t want to be left out of the potentially transformative impact of machine learning.
Digital transformation also suffers a skills shortage
Our survey showed that the skills shortage extends beyond esoteric areas such as machine learning and into the more prosaic skills involved in software development and basic infrastructure. As the world becomes more and more digital, organisations are looking to transform their underlying IT environment to better support what’s become known as the “API economy,” where the primary way for organisations to interact with customers and partners is through mobile and web applications, with face-to-face and even phone interactions becoming less and less frequent.
Digital transformation, for example, requires a trustworthy and secure infrastructure. With ransomware and state-sponsored hacking hitting the headlines, companies are also seeing cyber security as a strategic need in the increasingly-interconnected IT ecosystems that organisations are building with partners and customers. “Cyber Security Engineer” was seen as being a difficult-to-fill IT role by 14 percent of respondents in our survey.
Likewise, “digital transformation” requires greater organisational agility, both in software development and in IT infrastructure, and is another hard-to-find skill set. “Scrum Master/Agility Coach” was chosen by 12 percent as being difficult to hire, as was “Full Stack Architect.” Having an experienced scrum master available to coach employees can be an important factor in supporting the move to agile development. Likewise, as companies look to accelerate their digital transformation, having a good IT architect who can help transform and integrate legacy infrastructure into more modern settings is essential.
Current employees lack the necessary skills for this new world of technology
The problem with manpower is not limited to hiring the right people, either. One of the reasons companies are looking to hire outside is that their current IT staff lack the skills needed to move the business forward.
With the shift to the API economy, the mobile world has trained users to expect regular, if not daily, updates to an app. This requirement is straining the old waterfall approach to software development and deployment to the limit. As a result, many organisations are turning to DevOps -- an approach that links the development, testing and delivery of software into an integrated, continuous process. Yet DevOps is not easy to implement, and around 40 percent of IT decision makers in our survey named “experience with DevOps/continuous development” as one of the top three skills their organisation’s IT team lacked.
organisations are also clearly drowning in data and think their current IT staff isn’t quite keeping their heads above water. Around 33 percent of IT decision makers think their teams are lacking in “experience handling large-scale data and/or big data platforms.” Even basic “database administration and development” skills are cited as lacking by 32 percent of IT decision makers. This connects back to the difficulty that organisations are having in finding machine learning engineers, as it is big data that make machine learning applications so effective.
Testing and user experience (UX) are also seen as missing among current IT staff, with 31 percent reporting that their staff lacks “experience integrating UX testing tools/test automation frameworks.” This echoes the lack of DevOps skills, but it also speaks to the emerging impact that technology has on creating a good customer experience. If the primary interface between the customer and the organisation is through its web and mobile apps, then a focus on UX becomes of primary importance.