Setting the benchmark for data science in Asia

  • Centre of Applied Data Science partners US-based The Data Incubator
  • Producing data scientists who can communicate, not just compute
Setting the benchmark for data science in Asia

 
SHARALA Axryd (pic above) is a geek with a bold vision. The managing director of the Centre of Applied Data Science (CADS) aims for it to become the benchmark in data science education in Asia.
 
To achieve this, she has partnered with another geek – Dr Michael Li, executive director of The Data Incubator, to bring his data science curriculum to Asia.
 
The Data Incubator was founded in 2014 with grants from Cornell Tech, the computer science school at Cornell University, and is considered among the leading data science incubators in the United States.
 
Sharala credits Multimedia Development Corporation (MDeC) with connecting her to Li. MDeC is the government agency tasked with boosting ICT in the country, and has embarked on a number of big data and data science initiatives to develop Malaysia into a data science hub.
 
A former telco engineer with a Master’s in mobile and satellite communications, Sharala leveraged on her global telco experience to become an entrepreneur, focusing specifically on offering training to the telco industry.
 
And with telcos probably the biggest owners of customer data, it was no surprise that in 2014 big data and data scientists came under her radar.
 
This was also when she realised the scale of the problem.
 
“All the big companies know about the value they can extract from their data. The challenge was to find the right people, these data scientists, who can help them,” she says, speaking to Digital News Asia (DNA) in Kuala Lumpur.
 
“This is a problem they need help with, yesterday, so to speak,” she adds.
 
Speaking to IT vendors was not very helpful as they would typically advise companies that purchasing their products was the answer to big data challenges.
 
The industry apparently thought it needed statisticians or those with a statistics background.
 
“But when I spoke to these statisticians, I found they were very poor communicators,” says Sharala.
 
“They wouldn’t even look me in the eye. How were they going to be able to communicate what the data was telling them, to their colleagues?
 
“I then thought that the better solution lay within the big companies themselves. Many of them had business intelligence teams. Why not train them to become enterprise data scientists?” she adds.
 
The companies found the idea intriguing but wondered if Sharala could convert a domain expert into a data scientist within a short training programme.
 
This led to setting up CADS in January 2015, which Sharala describes as “our first-strike response.”
 
“It was specifically conceived to nurture a new generation of data professionals who can meet and exceed the needs of our digitally-disrupted world.
 
“We categorically do not believe in taking a shorter-term view of data analytic education, by focusing on churning out a lower standard of data professionals to meet immediate needs,” she declares.
 
Setting the bar
 

Setting the benchmark for data science in Asia

 
Her search for a world-class curriculum laid the seeds for the partnership with Li’s The Data Incubator.
 
“Over many phone calls and email exchanges, we discovered a shared commitment to the highest standards of data science education and decided to take the plunge together, with CADS granted the exclusive rights for South-East Asia for The Data Incubator’s curriculum,” says Sharala.
 
Two cohorts have graduated so far with 45 participants in total, 75% of whom were from the industry, with the remainder being PhD students or holders.
 
The focus on PhDs comes from how Li runs his programme, which is now offered in New York, the greater Washington area, San Francisco, and Kuala Lumpur.
 
The intake in the United States is very much focused on students who have PhDs, with the idea being that they would already have 90% of the background needed to be a data scientist.
 
The intake in the summer of 2015 saw 2% candidates admitted to the programme from the 3,000 who applied, according to Li.
 
“It is a very rigorous process and those who are admitted are then given the right tools to help them leverage on the open source ecosystem to learn how to use tool sets used by industry, rather than the tool sets taught in academia,” he explains.
 
Li has been a data scientist with Foursquare and Google, a scientist at NASA, and is a contributor on the subject of data science to the Harvard Business Review (HBR).
 
With his focus on PhDs, one would think Li is focused solely on brains, but surprisingly, the ability to communicate ranks as among the three core skills that he feels is critical to being a good data scientist.
 
It is a skill that he aims to bring out through his incubator, he tells DNA.
 
“Sure you need enough math and statistical knowledge, coupled with a programming background – but this needs to be matched with the ability to communicate and relate the data you are seeing to real-world business problems,” he says, having penned a specific article in HBR on the subject as well.
 
In essence, the curriculum from The Data Incubator is all about taking people with knowledge in each of the three areas – computational, statistical, and communications – and getting them to a level where they are strong in all three.
 
Sharala however has adapted the curriculum and intake to local needs and culture.
 
Firstly, recognising that she will not find 3,000 PhD graduates a year in Malaysia, the intake focuses mainly on participants from industry – and not necessarily from the tech industry.
 
In fact, the US experience has shown that the best data scientists are actually those with chemistry and biotechnology backgrounds.
 
In line with the strong industry participation, the programme is less about taking exams, although participants are encouraged to come in with a capstone project relevant to their company.
 
“Culturally, Malaysians tend to shy away when hearing about exams being part of a training programme,” says Sharala.
 
“The companies like this project approach as well, and we have moved away from a full-time course to a part-time course with three full days of lectures per week, with the course still being over eight weeks,” she adds.
 
Getting into the programme is difficult though, as Sharala is keeping to the very high standards set in the United States.
 
“But what we have done slightly different is to have some hand-holding in the form of extra classes, and this has had a huge impact for participants,” she claims.
 
“And yes, with the rigorous curriculum set by The Data Incubator, we are making the claim that our graduates can be called data scientists,” she adds.
 
Sharala is hoping the programme helps Malaysian companies first, and soon companies in South-East Asia, maximise value from their big data.
 
It will have to, if she is serious about setting the benchmark for data science education in Asia.
 
Related Stories:
 
A life in data, a handbook for data scientists
 
The data scientist market: Some thoughts and tips
 
Big data: Malaysia takes ‘small but significant’ first step
 
Addressing the data scientist glut
 
 
For more technology news and the latest updates, follow us on TwitterLinkedIn or Like us on Facebook.
 

 
Keyword(s) :
 
Author Name :
 

By commenting below, you agree to abide by our ground rules.

Subscribe to SNAP
Download Digerati50 2018-2019 PDF

Digerati50 2018-2019

Get and download a digital copy of Digerati50 2018-2019