Data talent crunch

Dutta is leading a team of about 20 data scientists who have been hired by Sembcorp as the group embarked on its transformation into a business at the forefront of changes in the urban and environmental landscape. In February last year, newly appointed CEO Neil McGregor laid out revival plans for the erstwhile conservative group as its earnings took a hit from the offshore and marine down cycle, as well as the declining performance of its utilities businesses in Singapore and India. McGregor said Sembcorp’s future lay in a focus on sustainability, which also entails repositioning the core Utilities unit to deal with the changing business of energy as it moves towards decentralisation, demand disruption, decarbonisation and digitalisation.

“When I heard about what Sembcorp was planning to do in the area of business transformation, bringing digitalisation as a key enabler, that was a very important message for me,” Dutta tells The Edge Singapore in an interview. “I felt there was a lot of innovative application opportunity.”

Before joining Sembcorp, Dutta had already worked for 12 years, with stints in the aerospace, mining and academic sectors. The common thread across those jobs was extracting insights from data, “whether you are solving a design challenge in the aerospace industry or a production efficiency problem in a mining company, or improving the carbon footprint in a power company”, he explains.

Still, Dutta was looking for a way to level up his skills in data analytics. “What I wanted to do for the next step of my career was not just to use data science as a day job but as a real innovative tool to make tangible and significant improvements to the business. We invest billions in high-value assets to produce power. How do we ensure we are getting the most out of these assets?” he points out.

And, Sembcorp’s turn to renewable energy was a big draw for him. “I have a passion to impact the environment, so when I saw that the opportunity would not just enable me to pursue my passion but also make a tangible difference to my employer’s vision and mission to impact the environment, it was a double [bonus].”

It might as well be rocket science

Data science, or what is essentially the marriage of statistics and computer science, has become increasingly sexy over the last decade or so, as advances in technology allowed people to sift through vast amounts of data to extract insights and, importantly, monetise or at least find business applications for them.

According to a study by the Asia Pacific Economic Cooperation, demand for data specialists is outstripping supply across Asia-Pacific. Tableau estimates that, in Singapore alone, 15,000 data specialists were needed to meet “base demand” last year. According to the company, there were about 9,300 such people employed in 2015.

“A key factor behind this demand is the urgency with which some businesses are looking to harness the exploding volume of their data,” says J Y Pook, senior vice-president, Asia-Pacific at Tableau, a software company.

What businesses typically seek in a “data scientist” is an individual schooled in the STEM (science, technology, engineering and math) skills. Other qualities to complement these skills are deep domain knowledge, business savvy and soft skills, including problem-solving and communication skills, creativity and the ability to work in a team. “These individuals are highly desired, but extremely hard to find,” says Randy Goh, managing director of data analytics company SAS Singapore.

“They need a thorough understanding of the domains in which their insights will be applied, including supply chain, finance, logistics and human resources. This comes from experience, which is difficult to teach in the classroom, making fully qualified data scientists hard to find,” Pook adds.

Goh notes that the turnover of data professionals tends to be relatively high. People stay in the job for an average of 1½ to two years. “[This] adds to the challenge of an already limited pool of talent,” he says.

Moreover, the role and requirements of data scientists are constantly evolving, Tableau’s Pook notes, which in turn puts pressure on the existing pool of data scientists and those entering the workforce.

“The hiring crunch is felt more acutely by organisations without sufficient data assets and training capabilities to grow talent internally,” says Mohan Jayaraman, regional managing director of Decision Analytics and Business Information Services at data services company Experian Asia Pacific.

Dutta observes that while the company has been able to fill junior positions from the local talent pool, it has been more challenging for the more senior positions. Most of the applicants for such roles have tended to be foreigners.

Plugging the shortfall

Against this backdrop, the government and tertiary education institutions in Singapore have launched various programmes aimed at addressing the gap between demand and available talent.

At the National University of Singapore, the School of Computing launched its Bachelor of Science (Business Analytics) six years ago. It has since opened up its data science and business analytics courses to students outside of its school, and has seen its intake for the degree programme increase from 50 students in the first intake to about 150 students in the current academic year.

“This degree programme has received tremendous interest over the last few years as more prospective students who were originally interested in degrees such business administration and law are realising the importance and value of getting a technically oriented education in data analytics methods and techniques applied in the business domain,” says Goh Khim Yong, vice-dean, School of Computing, NUS.

He adds that NUS has a multi-disciplinary programme that “provides a solid foundation of technical computing knowledge and skills, data analytics methods and techniques, as well as business domain knowledge — all of which form the backbone of the data science discipline”.

Students have to undergo either a six-month internship or a 16-month work-and-study programme with companies. NUS also offers data analytics courses for working professionals.

Meanwhile, the Tech Skill Accelerator, run by the Infocomm Media Development Authority (IMDA) of Singapore, equips jobseekers with relevant skills through schemes such as company-led training, immersion, placement and mid-career conversion programmes.

One of the accelerator’s beneficiaries was Sheralda Chen, a data analyst at media, digital marketing and communications group Dentsu Aegis Network. Chen had developed an interest in data analytics while an undergraduate at NUS Business School. “In my last year of university, it became apparent that data-driven marketing was the road forward. At that point, I was neither comfortable with nor adept in numbers, much less big data,” she says.

The accelerator provided an opportunity for her to be part of a data and analytics programme jointly offered by IMDA and Google. “It seemed like the perfect opportunity to gain the skills I desired and a doorway to well-known players in the industry,” says Chen.

Indeed, what is key is real-world experience, according to Experian’s Jayaraman. “To meet the growing demand, educators are starting to roll out various courses that are focused on data science, analytics and statistics,” he says. “However, deeper collaboration between stakeholders, including enterprises, local start-ups, academic institutions and government agencies, will be crucial to the development of skilled talent that possesses hands-on experience.”

Chen’s experience has helped her company take its business to the next level. “With data analytics, we can push the boundaries of performance for our clients by looking beyond the usual performance indicators,” she explains. “It has also alleviated repetitive and menial tasks for the team and brought massive time savings. This has allowed them to focus more time on the work that truly adds value for clients.”

There could be other ways to plug the shortfall in the number of data scientists available here. Companies could leverage data tools while building up their own skills. SAS’ Goh lauds the measures taken by the government in the recent Singapore Budget to provide funds to companies to invest in developing their employees’ digital skills, although he acknowledges that it will take a concerted effort by the government and all segments of society.

“What will ease off the pressure of the shortage of data scientists is fundamentally changing the usage of data within businesses — from the siloed usage within Business Intelligence and IT teams to empowering more people with data across the business,” says Tableau’s Pook.

The combined efforts of schools and businesses can help address the data talent crunch, says Experian’s Jayaraman. However, “to remain competitive, the workforce will have to continually reskill and upskill”, he adds.

Pook believes engaging younger students will be key in making up the shortfall. “This is a good start, but to have a real impact, we need to engage students with data at a younger age. There are various programmes to expose young people to data early on,” he says. “One-off initiatives are not enough, and they need to be more widespread. Just as companies are using data in many different functions, from sales to marketing and HR, schools need to bring data into the different subjects so that children can learn to uncover the hidden insights and communicate them effectively.”

Ultimately, it may very well be the challenge and versatility of the job that would attract fresh talent into the sector. For Sembcorp’s Dutta, it was a special course on artificial intelligence in his final year of university that sparked his interest. “That changed my whole perspective on technology, and I made the conscious decision to move full time from engineering to computer science, specialising in AI,” he says. He went on to obtain a master’s degree and doctorate in AI.

Dutta’s work revolves around solving the various business and operational challenges that Sembcorp faces. This includes improving the efficiency of assets, ensuring the work environment remains safe and building predictive models for trading and selling decisions. Dutta and his team also work on price forecasting for Sembcorp’s budding merchant business. “We are trying to capture all these market drivers and create prediction models, using machine learning and AI, to understand where the market is heading. [Then, we give] that insight back to our merchant, retail and commercial teams to make smarter decisions,” says Dutta.

“It’s not just about solving business challenges, building new applications or using machine learning and [AI] but also about the opportunity to understand a whole new vertical, a whole new set of business challenges and processes,” he adds. “That is the fascinating part of being a data scientist.”