While businesses slowed for most companies amid the Covid-19 pandemic, ed-tech company upGrad has been seeing a surge in learners enrolling on its platform. And just as it widens its offerings of industry-relevant data science programs to meet growing demand, the company has also been onboarding some of its alumni into its employee base.
In fact, the company is in the process of scaling up its data science team, and are looking at filling some senior and mid-level positions in the coming months.
“We want to make our data science team stronger, and will be looking to onboard enough people to make it into a 15-20 member team in the next couple of years,” says Mayank Kumar, co-founder and MD at upGrad.
How Important Is Education To An Education-Focused Company?
“The educational background of a candidate plays a significant role in the overall hiring process, and with an alumni base of over 10,000, we have a good and dependable source to track their competencies,” says Mayank.
However, he is quick to clarify that although education is a differentiator, it is not an eliminator when it comes to hiring for data science positions. Furthermore, the skill sets that he seeks in candidates largely depends on the kind of profile they are recruited for.
“Within the field of data science, there can be multiple roles like data engineers, data analysts, or data scientists, and at upGrad, we have open requirements for all of them,” says Mayank. “Additionally, the skill sets sought vary based on the level of seniority in the same role,” he adds.
ALSO READ: The Dos & Don’ts Of Hiring Data Scientists
At upGrad, a candidate is tested for their problem-solving skills, hands-on experience with coding, ability to comprehend from research and available resources, and their capacity to put them into applications.
“In addition to these, we also look for business acumen and leadership skills, especially when recruiting for senior positions,” says Mayank. “So we hire a mix of both highly-skilled candidates as well as execution-level ones who can look at solving problems that range from low to medium to high complexities in the next five-year run,” he adds.
Hiring Process for Data Scientists at upGrad
The hiring process at upGrad is spread across four parts, each of which varies depending on the experience level of the position. “A senior-level position, where we expect 8-10 years of work experience, will take longer to convert than a position at a junior level,” says Mayank.
In the following paragraphs, we take a look at these four parts that tie together the hiring process for data scientists at upGrad:-
Broad-Fit & Expectation-Setting Round – The hiring process starts with a discussion round with the Product Head or Technology Head to establish both candidate’s and employer’s expectations.
“At this point, we brief candidates about upGrad and give them an overview of what the prospective job role holds for them,” says Mayank. “This conversational round also leads us to understand the candidate’s breadth and depth of experience, including what are the kinds of domains they have worked in, the range of problems they have solved, etc.,” he adds.
As part of the same round, the company also shares certain problem scenarios with candidates, which they are expected to solve using their technical acumen and method. They are then judged based on their process and coherent approach of solving problems, rather than based on which algorithm they use.
Deeper Technical & Evaluation Round – At this stage, data science experts in the company sit on a video call with candidates with screen sharing, which would involve testing their coding skills and understanding. This round involves written or practical assignments, which can last from 90 minutes to up to 2 hours.
Leadership Round – At this point, shortlisted candidates interact with senior leadership for an in-depth analysis of capabilities and skills. Based on their level of seniority, they may have to interact with one or more senior leaders.
HR Round – This is the last and final round for checking whether or not the candidate is an ideal fit at upGrad. Here, the HR team checks for organisational alignment and cultural fitment.
Top Interview Questions For Data Science Positions At upGrad
- What are the differences between supervised and unsupervised learning?
- State an example when you have used logistic regression recently.
- What is Random Forest? How does it work, and where would you use it?
- What do you understand by explainability in a model?
- What ranking algorithms are you familiar with?
- How do you communicate what you do to a non-technical audience? Given the R&D nature of your work, how do you set expectations with other stakeholders who may not understand data science?
- Outline an approach for solving a sample problem. Example: How would you come up with a solution to identify plagiarism?
- How do you decide how much data you need for exploration, training, and testing? Give examples from your past experience.
- What are the various domains to which you have applied data science? Provide a summary.
- What has been the most challenging project you’ve worked on? What was difficult about it?
- How do you keep yourself current with the latest research and state of data science? What magazines, sites, and people do you like to follow?
- What kind of challenges have you faced in operationalising the models you develop?
- What is the biggest data set that you processed and what issues did you face in processing it?
- What are the ‘best practices’ in data science that you like to follow?
Sourcing Data Science Candidates
According to Mayank, a lot of skilled data scientists try to get scores on various hackathon platforms, which have emerged as the non-traditional ways of getting hired or shortlisted.
“It’s a prevalent culture in this domain, and platforms like Kaggle allow individuals to submit codes and get points, and they have become relevant sources to acquire talents,” he says. “We have also tried hackathons in the past and have shortlisted candidates who have done well,” he adds.
That is not to say that traditional modes of hiring are ignored. Reliable references from portals like Naukri or LinkedIn, as well as those from existing employees, help speed up the hiring process at upGrad. The company also encourages internal job transfers wherever possible.
What are the challenges faced by the company when recruiting for its data science positions?
“Lack of knowledge about advanced statistical tools and other technical methods that can help professionals make optimum use of data sets poses a big challenge and has been contributing to the shortage of data scientists in the industry,” says Mayank. “The skills are getting redundant, thereby increasing the data science skills gap,” he adds.
However, according to him, the Covid-19 outbreak has introduced a wave of automation for businesses and has given existing data scientists the opportunity to upskill and reskill themselves.
“There’s a dual challenge in hiring for data science roles,” explains Mayank. “It’s not just about finding the right candidate, but also the capability to evaluate them effectively. For instance, if one is hiring for a principal data scientist position, and there’s no senior data scientist in the team, it’s difficult to get a full assessment of the candidate’s abilities,” he adds.