Instead, focus on hiring the technical skills needed to build the team, and the soft skills needed to work on the team
The data science unicorn is a somewhat mythical person who is a leader in data science, technology, and business. Of course, these candidates practically don’t exist, nor do they necessarily make strong team members. As data science teams have grown, businesses have moved away from trying to find that one person to fill different roles; instead, companies have realized the benefits of hiring employees with specialized, complementary skills.
Data scientists are still in high demand. In fact, data science job postings have increased by 256% since December 2013. It seems that no industry is immune to this data scientist shortage, as global companies continually seek qualified talent. And luckily, this search has become more niche. While many job postings still say “data scientist,” the descriptions are beginning to lean towards more specialized needs. Furthermore, companies are seeking candidates with strong soft skills that can complement a data science team.
“It’s true that having advanced knowledge of mathematics and programming is fantastic background for a data scientist,” he says. “But, in any company, you won’t find just one data scientist doing it all—just like Michael Jordan couldn’t have scored so many points without Scotty Pippen at his side, data scientists all bring their own skills to the table that together build an ideal team.”
The new data science team
The specialties within data science are numerous and growing. From data mining and statistical analysis to deep learning and cloud computing, data scientists have options when it comes to choosing where to focus. And companies have options when scaling a data science team.
When building a team, it’s critical to understand how the project will be introduced, maintained, and scaled not only in terms of technology but also in terms of individual roles within the project. Some companies opt to build a team using an IT-centric structure. Those that use this option typically utilize some sort of machine-learning-as-a-service (MLaaS) software that has low barriers to entry. The entirety of the project, including data preparation, model training, interface creation, and deployment all happen within the IT infrastructure led by the IT team.
Another option is using an integrated structure. Here, the data science team prepares the data and trains the models, and the IT team then takes over to evaluate and deploy the models. This approach requires a robust data science team with complementary skill sets. A third option is to run the entire process, from data preparation to deployment, within a dedicated data science team. In this scenario, the data science team must have IT infrastructure knowledge and skills to get the models to deployment.
Often, companies build a team that incorporates data scientists, IT employees, and business analysts, or people that can translate findings to business value. Open communication between these three teams working as one reduces the risk of models not being deployed or improper infrastructure issues. Of course, building the data science part of the team can be difficult with the current skills gap. And with that scramble to find talent, new roles and specialties are emerging in the data science world.
In-demand data science specialties
While it has always been difficult to fill roles such as Chief Data Officer or even an experienced deep learning engineer, there are new positions that have risen in the last few years that companies want to fill.
More companies are applying data-driven methods to cybersecurity in order to improve the prevention of breaches. Whereas security approaches tend to be applied post-breach, data science has been helping companies tackle cybersecurity both pre- and post-breach. Using supervised machine learning, this technology can find behavioral features from executables run in isolated environments, which makes it harder for malware authors to go unnoticed. Anomaly detection post-breach means that attackers who already are inside the system can be found using self-learning models that understand normal behavior. With the surge of cybersecurity hacks and their rising costs to enterprises, it’s no wonder that companies are recruiting data scientists that specialize in security.
Another in-demand position in the data science world is the financial data scientist. Finance professionals have long been performing data science tasks such as risk assessment and forecasting. Data science helps improve and automate many of these tasks. A candidate that understands finance and also has a robust knowledge of data analysis, programming, and statistical techniques becomes a financial data scientist that can dramatically improve company performance.
Don’t try to find the unicorn
Companies ramping up a data science team shouldn’t waste time trying to find the unicorn, because it will slow down hiring drastically. Instead, they should focus on the technical skills necessary to build the team and the soft skills needed to work on the team. In my opinion, large enterprises will likely continue to look for security and finance specialists, as more candidates are likely to get training in these areas to fill this need.
While artificial intelligence and machine learning can revolutionize a business, these technologies cannot do much without proper model building and infrastructure in place, making it worthwhile to slow down in order to scale correctly.