Software developer Sufyan Khot’s LinkedIn post titled ‘I quit data science’ sparked a lively debate on the platform. At the time of writing this article, the post had garnered close to 1,400 ‘likes’ and over 90 comments–a testament to how the post struck a chord with a large number of people.
Quitting data science
“Harvard Business Review hailed Data Scientist as the sexiest job of the 21st century. LinkedIn and Glassdoor continue to rank data science as one of the top professions. The median experience of people in the profession is probably less than two years, but the power it wields over the industry is huge. An average data scientist salary in the US is roughly $117 – 120K, much higher than what an experienced software developer might have. So these are two powerful metrics in terms of data science as a lucrative career from a youth’s perspective,” said Shekar Murthy, Senior VP, Presales, Solution & Professional Services, Yellow.ai said in an earlier interview with Analytics India Magazine.
That said, data science is not for everyone. Khot realised it the hard way when he quit his regular job as a software engineer to pursue data science. “At my first job in 2019, I saw a team of data scientists at my organisation. I was very intrigued by the job overall and felt I had the aptitude to be one of them given my interest in mathematics, programming, and statistics. I zeroed in on a six-month course in data science because of which I had to quit my job.”
Khot did exceedingly well in both theory and practicals. However, at the fag end of his course, he knew it’s not his cup of tea. “There were two main reasons for this decision. Firstly, a large part of a data scientist’s job is quite monotonous, especially cleaning and processing raw data. A few estimates suggest that a data scientist spends as much as 80 percent of his/her time doing that. Secondly, despite reports of companies just waiting out there to hire data scientists was not true in my case. I felt that good jobs were far and few in between.” Khot also warns aspirants to not fall for the glamour element of the job.
Now, Khot is back to his old line of work.
From recruiter’s perspective
Harsh Gupta worked as a data scientist at prestigious organisations such as WWF and John Hopkins for six years. Gupta is currently the founder and CEO of Oklahoma-based organisation ProtoAutoML, an autoML software provider. Gupta shares how he was a 20-year-old graduate who spent the first six months being a data scientist carrying monotonous data cleaning and processing tasks. “In those six months, I was required to make exactly one regression model,” he said.
“Many companies do not have the proper machine learning tools and/or still rely on legacy systems. A data scientist entrant is probably coming from an academic world and may already be exposed to platforms like Kaggle, GitHub, and other open-source projects. So they may come with some unrealistic expectations. They would want to straight away work on high-end projects, while in reality a large part of their time will be spent in making sense of the data,” said Gupta.
He feels companies are also at fault as they often fail at clearly defining the job role being hired for. For example, he said companies use buzzwords like AI, machine learning, and data science while advertising for a job but, in reality, may expect their employees to also work on the Tableau and business intelligence side. He said clerical jobs related to preparing the data can be easily automated so that data scientists can work on more skill-based processes.
For his organisation, Gupta said, he makes the job responsibilities clear from the start. “I only hire people who have had considerable experience in working with data science projects and have a GitHub of their own. This tells a lot about the candidate’s experience with data science, making it easy for both the employee and employer to set the right expectations,” he said.