AI has long ceased to be a subject discussed only by deep tech experts. It’s now a part of our everyday lives, affecting not only how we work and do business, but also how we shop, bank, and get around.
That said, just because the tech has arrived doesn’t mean that its story is complete. Its use cases continue to grow, especially in Southeast Asia, a developing region that faces issues such as a lack of financial inclusion, traffic congestion, as well as poor infrastructure.
A report from McKinsey & Company outlines even more ways that AI can contribute to positive social outcomes in Southeast Asia, helping sectors like education and remote healthcare as well as in the development of medicine.
The potential for AI-enabled progress is immense.
Grab is driving progress
Southeast Asian everyday super app Grab plays a big role in furthering the application of the tech.
“We believe that AI can unlock new ways to change the lives of millions of Southeast Asians for the better,” says Foo Wui Ngiap, the firm’s head of AI initiatives.
The regional unicorn is known for its consumer-facing services, such as ride-hailing, food delivery, and mobile payments. However, not many people realize just how much of these platforms are a product of effective AI implementation.
“A lot of what you see in the Grab app today is powered by some form of AI,” says Foo.
For instance, a user of Grab’s ride-hailing service will not simply be assigned to the nearest driver when they request for a ride. “The driver closest to you may be about to wrap up his day. Your destination is out of his way, and he’s likely to cancel on you,” explains Foo. He adds that the company takes into account “over 50 attributes – driver profile, the timing of his last ride, location, time of day, etc. – to find you the best match.”
In fact, a feature that Grab rolled out last year uses AI to allow drivers to complete as many jobs as possible between their current location and their homes, Foo tells Tech in Asia.
This kind of detail is present in each service the company offers. For example, it uses natural language processing (NLP) techniques to address customer feedback at scale and help users discover their desired services quickly. It also analyzes behavioral and earnings data to extend lines of credit to its driver partners that enables them to cover daily expenses like fuel and utility bills.
Grab does this with “hundreds of thousands of matches and millions of transactions at any second,” Foo says.
A bigger purpose
However, changing lives goes beyond just providing great service.
“We want to use AI to solve Southeast Asia’s biggest problems, to find ways to elevate the quality of life for the 650 million people who live here,” Foo emphasizes.
To this end, Grab has four strategic focus areas for AI over the next 12 months, which includes fixing traffic congestion issues and building a richer understanding of the region’s roads, solving fraud to protect users across the Grab ecosystem and making transport safer, increasing personalization and marketplace optimization, and bridging Southeast Asia’s communication gaps.
The company’s business objectives would work hand in hand with these goals. For example, the road network data that the company collects from its driver partners – who collectively traverse 320 million kilometers each day – is used to improve its maps and can also be used to reduce traffic congestion. In this way, Grab has discovered more than 750 km of undocumented road segments in Jakarta alone.
Effective optimization of demand and supply doesn’t just mean faster delivery times. It could also lead to fewer cars on the road and less food waste.
AI will also support efforts to ensure the safety of passengers, working in concert with real-time safety monitoring to detect anomalous signals such as route deviations. “We’ll continue to invest heavily in safety so that we get to zero preventable incidents,” adds Foo.
As for bridging communication gaps, he points to the diversity of languages across the region, and how they aren’t as well-served by the large providers of translation services. Grab is actively working with AI partners to address this issue.
Foo is confident that the company can drive these positive changes. “Firstly, with the data we have – every day we generate about 40 terabytes of data – no one understands the region better than we do,” he claims. “Secondly, we are committed to developing new tools and technologies that are specifically tailored to the nuances of the region.”
“Ultimately, we want to make good use of this data – either to power intelligent and delightful customer outcomes or for the good of Southeast Asia in a bigger way.”
Image credit: Grab
This year, Grab has invested over US$100 million in AI development, spread across infrastructure, talent, technology, and strategic partnerships. The company is expecting to grow this amount by 50% in 2020 as it moves toward becoming – as Foo puts it – an “AI-everywhere” operation.
“We want to build an organization that thinks and breathes AI,” he says.
Internally, this involves getting all of its roughly 2,000 tech and engineering employees to undergo training in artificial intelligence, as well as embedding members of its 300-strong AI and data science team – one of the largest in Southeast Asia – across the organization. “We want everyone equipped to actively think of ways to leverage AI in their jobs,” Foo explains. “Put it this way: AI in lots of little things makes the big thing better.”
Externally, the company has partnered with Microsoft, tapping into the US tech giant’s capabilities in facial recognition to enhance verification features for both drivers and customers, and also in NLP to improve its sentiment analysis abilities, among other features.
Grab is also working closely with universities to establish laboratories, mentorship programs, and internships to build the AI and data science talent pool in Southeast Asia. Foo notes that this is key if AI is to reach its full potential in the region.
“Data scientists are in short supply even in the global AI hubs, and that shortage is felt even more acutely in ASEAN,” he says. “We set up local R&D centers in cities like Singapore, Jakarta, Ho Chi Minh City, and Kuala Lumpur as we see significant potential to grow our local talent over time. It’s a long-term investment back into the region.”
While it’s easy to get lost in the many ways to tackle AI adoption, the company is laser-focused on the ultimate objective.
“We’ve never built tech for the sake of it, and that applies to our AI strategy as well,” Foo stresses. “AI isn’t the end goal. Solving problems is.”
-Tech in Asia