Data science is a growing field, but hiring the right people can be challenging. Here, HackerRank’s Vivek Ravisankar offers his advice.
In LinkedIn’s recent Emerging Jobs Report, data science was one of the fields predicted to grow significantly in 2020. As with any expanding industry, you might expect open roles to be filling up at a rapid pace.
But in data science, the recruitment process is actually being slowed down, according to HackerRank. The tech-focused recruitment site says it may be due to hiring approaches being unable to deal with the different specialist skills that applicants might have and companies might need.
The US company was founded by Vivek Ravisankar and Harishankaran Karunanidhi to provide clients with tools that could help them navigate the data science recruitment process more efficiently, with features such as role-specific technical assessments.
Ravisankar, HackerRank’s CEO, told us more about the problems that he and his team have identified in the market and why it’s important that we solve them.
‘Data scientists must know how to communicate with their colleagues to get to the question behind the question’
– VIVEK RAVISANKAR
Why is it important that we get better at understanding how to hire data scientists?
The main reason to get data science hiring right is that doing so can have significant business impact, ranging from identification of areas where you should double down to the addition of product functionalities that become stronger with growth.
Unfortunately, there’s confusion in the market about the different roles in data science, such as between data scientists, data engineers and AI engineers, for example. Clarifying exactly what those roles entail and what skills are required for each is an important starting point.
Is there a shortage of data scientists with the right skills at the moment?
Demand for data scientists is high: it has grown 256pc [in the US] since 2013, and our 2020 Developer Skills Report found that data scientists are the top hiring priority for nearly one in six hiring managers globally.
However, 55pc of data scientists come from non-computer science backgrounds, including physics, maths and biology.
Companies need to ensure they define exactly what they’re looking for when designing the data science hiring process. This will lead them to candidates with the right skills much faster.
What are the main attributes to look for when hiring data scientists?
The best data scientists can build effective models, use appropriate techniques for different kinds of problems and strategise on augmenting data sets – maintaining clean, extensive data sets is the biggest challenge in many data science projects.
They are also excellent communicators with business acumen and a ‘boardroom presence’, and can build strong teams to support them.
Are there any alternative attributes that help someone to stand out?
Because data scientists are required to work cross-functionally across organisations, collecting and analysing data from a broad set of data sources, they must know how to communicate with their colleagues to get to the question behind the question.
The ability to articulate exactly what needs to happen based on the data model that is built is a key trait for a data science leader. Data can be highly subjective depending on how it’s sliced and the more complex it becomes, the harder it can be to derive meaning.
Is there anything in the hiring process that could be improved?
The two main improvements needed are standardisation of the hiring process and clear differentiation between each data science role.
The diversity of data scientists’ backgrounds – nearly 60pc learn their skills outside a university, and about 55pc come from backgrounds including physics, maths and biology – makes it tough for hiring managers to standardise the recruiting process and know which skills to prioritise in their testing.
How could that be achieved?
Hiring managers can start by seeking candidates beyond those with traditional software development backgrounds, but they also need tools that standardise the hiring process, simulate real-world job environments and assess skills specific to data scientists.