Photo - Stephen Brobst, Chief Technology Officer, Teradata Corporation
One of the key messages from data specialist Teradata to Malaysia is that a strong Big Data Analytics (BDA) culture is one that celebrates failure as being a key part of the path to successfully generating real value from analytics.
Speaking to media in Kuala Lumpur recently, Teradata's global chief technology officer Stephen Brobst and new Malaysia country manager Saqib Sabah said data scientists are in short supply globally though some countries are more actively fostering both the scientific and soft skill sets necessary.
"Data scientists need to have a scientific skillset, which includes creative thinking ability coupled with innate curiosity and the right talent does not necessarily have to come from an IT background," said Brobst.
"Hard sciences such as physics, astronomy and so forth tends to develop the right blend of skills and personalities," he said.
"However, another major factor is for countries to build the type of culture that really fosters BDA talent, which needs to be given a certain measure of independence and space to pursue innovation without fear of failure," Brobst added.
"In Silicon Valley, organised through a LinkedIn group, failure is celebrated and shared in regular meetings, by data science professionals as an essential part of being successful," he said, adding that Malaysia's national ICT agency MDEC's BDA community was a positive initiative.
Teradata in Malaysia
Teradata veteran Sabah, who recently replaced the late Craig Morrison as company's Malaysia lead, said Teradata in Malaysia will continue and further build on the strategies already in place, which included helping the Malaysian government realise its ambitions to become a regional BDA hub.
Photo - Saqib Sabah, Country Manager, Teradata Malaysia.
"Working mainly with MDEC (Malaysia Digital Economy Corporation) and various educational institutions, Teradata is already helping to develop BDA talent," he said. "We consider our cooperation with MDEC and local universities as vital to build a talent pool, which can support different industries."
"In addition, we will carry on with social analytics projects such as tracking dengue and healthcare in order to nurture and identify local talent," Sabah said.
"You need data professionals who are have business awareness as well as being tech savvy," added Brobst. "Students with hard science backgrounds - including physics, social sciences, maths and statistics - rather than computer sciences will know how to question data and open up possibilities."
"Problem solvers are drawn to cultures that give them the space and freedom to work with data reservoirs," he said. "Failing gracefully and proudly learning from failure is part of Silicon Valley's innovation culture."
"Malaysia can build such a community of data scientists and company leaders [such as MDEC's Big Community]," said Brobst, who gave examples of developed countries such as Germany, which has the technical skills but has not yet built a culture that sees failure as productive.
Both Sabah and Brobst added that BDA adoption in Malaysia was still in the "infancy stage" similar to last year.
"However, the telco and financial services as sectors in Malaysia are in the vanguard of BDA adoption," said Sabah when asked about local outlook. "Recent price and product wars in the Malaysian telecoms space demonstrate how BDA is being used in the intense competition."
"From a technological aspect, Malaysia is ready but are people culturally ready yet is the question," added Brobst.
Grain of sand
In an earlier forum with data analytics professionals from multiple sectors in Kuala Lumpur, Brobst outlined the latest developments in enterprise BDA, which included a presentation - Sensors Everywhere and the Internet of Everything covering the economics of data.
"Moore's Law has allowed economics to make possible the Internet of everything as everything can be sensor-enabled," he said. "The number of sensor enabled devices will increase from 9 billion to 25 billion devices in the next five years. Pets can be chipped not just things."
"By using sensors, we can measure everything that's going on in the network including vehicles, supply chains, and even the human body for healthcare," Brobst said.
He talked through some example, one of which was healthcare. "A research project between Google and Novartis is to make a contact lens to measure blood sugar levels in the fluid around the eye. A supply chain example, a well-known fast food company uses sensors to monitor temperatures in trucks carrying food supplies."
"Transistors are currently 14nm (nanometres) in size and be 8nm by 2019," said Brobst. "There's some research and development to be done but we can't predict what will be developed beyond 2021."
"Today's sensor is already the size of a grain of sand and costs less than a jellybean," he said. "In addition, people are already sensor enabled through cellphones. Sensors are becoming more cost effective, cheaper, faster, and smaller as we move towards nanotechnology."
Bottom line for Enterprise
What does this mean for enterprise analytics? Brobst outlined new sources of data will demand new provisioning models.
"We must widen our thinking," he said. "A database does not have to be relational. Data can be stored like a graph for examining data such as behaviour and tracking money laundering."
Sign up for CIO Asia eNewsletters.