Why is it so Hard to Find Good Automation & AI Talent?
Technology always evolves at “warp” speed, but in the past few years there has been an explosion of new technologies such as AR/VR, Blockchain, IOT, RPA Automation, AI, etc. Even though most of these technologies still have yet to go through the rapid evolution and adoption cycles, there is no shortage of experts talking about companies that are getting left behind if they don’t adopt these technologies quickly. In reality, objectivity must prevail when evaluating which technologies to invest in and which of them offer the best ROI. For this article we will focus on the Automation and AI trends which have real, tactical, and strategic benefits for businesses, and how to find the best Automation & AI talent to implement these technologies.
Many firms constantly face the challenge of “re-skilling” their existing workforce. For e.g. training a legacy COBOL or .NET Developer and re-skilling them with new age technologies. The issue with “re-skilling” is that only a fraction of the workforce makes a successful transition to the new age technologies. Many COBOL Developers found it hard to move into GUI Based Development tools in the late 90s, just as many Microsoft and Java based Developers found it hard to transition into the rapidly evolving technology frameworks such as Angular, React, Scala, and so on. Another reason there is such a challenging transition is that legacy code bases have to be maintained, and many such Developers are stuck in those maintenance activities.
AI based technologies, in many ways, change the classic developer paradigms. In AI, even though the developers still need to understand Software Development, they also have to transition themselves into Data Scientists in most cases. AI is not about input and output type programming, it is about taking data in and predicting the outcomes. Not all the Developers will be developing AI models, but many of them will be training and consuming them in their applications. Hence, without the proper academic background in Math, Statistics, or Computer Science, very few will be able to make a successful transition into Data Science and AI.
We also need to differentiate Machine Learning Engineers from Data Scientists. Machine Learning Engineers can be made from traditional programmers that can be trained to understand different types of ML models and consume them in traditional programming. For e.g. a Microsoft .NET developer can consume machine learning libraries and APIs in their software to make the software smarter. However, in many cases, one needs to understand the inner workings of the machine learning model, and without the academic background and/or data science training and aptitude, it’s hard for one to solve an AI problem.
Automation technologies such as Robotic Process Automation (RPA), on the other hand, are relatively easier to grasp. However, as easy as RPA may sound, just as finding good QA Automation developers was daunting, it is also very hard to find folks who have the right combination of technical and business skills to be able to map processes for automation and deliver an automation program that can truly scale. Automation, in many ways, is a logical stepping stone to AI because the many process and workflow problems encountered eventually deal with semi-structured and unstructured data which require some sort of AI algorithm to sort through it.
At Accelirate, we consistently recruit from U.S. Colleges and Universities that are developing advanced Mathematics, Statistics, and Computer Science programs. We take pride in hiring team members that have the right foundations upon which we can build their careers further; and although some of our RPA Engineers come from traditional QA Automation and Software Development, we seek to aggressively induct Data Science graduates into our training programs. We believe that they have the appropriate background that allows them to create autonomous bots that can function far beyond the capability of RPA Bots.