by James Henderson

The smart data driving Grab’s agenda in Southeast Asia

Aug 21, 2019
Artificial IntelligenceBig DataStartups

Mark Porter, CTO of Transport, Mobility and Core Technologies at Grab, outlines how the ride-hailing unicorn is leveraging artificial intelligence and machine learning to harness the daily accumulation of customer data

Mark Porter, CTO of Transport, Mobility and Core Technologies at Grab
Credit: Grab

Grab stands tall as the ride-hailing giant of Southeast Asia, the technology darling that transformed into a unicorn.

A startup with a story – legend has it, on a street, waiting for a taxi – and an enviable platform to match, facilitating over 2.5 billion rides across the region since launching in 2012.

This is the story of a lifetime for a blossoming business dominating the markets of Singapore, Malaysia, Indonesia and Thailand, in addition to Vietnam, the Philippines, Cambodia and Myanmar.

But such explosive growth at scale creates challenges, specifically around processing data – approximately 20TB on a daily basis – and harnessing relevant information to improve internal and external decision making. Being a treasure trove of knowledge is one thing, maximising its potential is another.

“You’ve heard people say that data is the new oil, currency or gold,” said Mark Porter, CTO of Transport, Mobility and Core Technologies at Grab. “And certainly, in my 30 years in technology, I’ve seen the power of data to develop and improve services for customers and transform the way businesses operate.

“With each new service we launch or city we expand into, the levels of complexity in Grab’s data grows almost exponentially. Each new data set adds a new dimension to explore and enables us to derive even more insights.”

Consequently, Porter, who was appointed to the role in October 2018, thinks that a key priority for the ride-hailing giant centers around the “smart” use of data – a common challenge facing any business undergoing rapid expansion.

“We want to minimise time and effort as we explore and experiment with the data, while maximising the meaningful and impactful insights we gather so that we can further improve the experiences we deliver to our customers,” explained Porter, when speaking to CIO ASEAN. “Going hand in hand with that is our growing investment into artificial intelligence [AI].”

AI advancements

AI, enhanced through machine learning, is emerging as a force for good within organisations of all sizes and sectors, as market spending edges towards US$35 billion in 2019, representing an annual increase of 44 percent, according to IDC.

“While organisations see continuing challenges with staffing, data and other issues deploying AI solutions, they are finding that they can help to significantly improve the bottom line of their enterprises by reducing costs, improving revenue, and providing better, faster access to information thereby improving decision making,” said David Schubmehl, research director at IDC.

As illustrated by Schubmehl, the AI use cases expected to see the most global investment this year are automated customer service agents (US$4.5 billion), sales process recommendation and automation (US$2.7 billion) and automated threat intelligence and prevention systems (US$2.7 billion).

Delving deeper, five other use cases forecast to experience spending levels greater than US$2 billion in 2019 include automated preventative maintenance, diagnosis and treatment systems, fraud analysis and investigation, intelligent process automation and programme advisors and recommendation systems.

But despite a clear direction of AI travel in the market, Seattle-based Porter insisted Grab is “only just scratching the surface” in terms of utilisation.

“I think AI will disrupt business and industry in ways that are hard to anticipate,” he outlined. “At a very fundamental level, AI allows us to do things faster and smarter than before.

“For example, the desire to understand customers and create products which maps to their needs hasn’t changed much over time. What’s changed is that we can now use machine learning to build models to predict ‘soft’ behaviours like consumer preferences in days and deploy software that adjusts to those preferences in weeks.”

Porter joined Grab after five years at Amazon, where he was most recently general manager of Amazon RDS, Amazon Aurora and Amazon RDS for PostgreSQL.

For the industry executive – who also draws on technology experience at NASA, Oracle and Caltech – the introduction of AI and machine learning represents a “step-change” in running a business, citing its emergence as “disruptive” in nature.

“It is disruptive because if your competitors are doing it and you’re not, customers will naturally and justifiably move to consume products and services which are more attuned to their needs,” he qualified. “In our products, we’re using AI and machine learning for things as varied as improving job allocation efficiency to surfacing food recommendations, monitoring passenger safety and catching fraud.”

Innovation at play for Grab includes how the business leverages AI and machine learning to enhance safety through its Driver Fatigue feature, designed to calculate a ‘fatigue score’ based on factors such as how long the driver has been on the road, time of day, rest between shifts and even age and profile.

“When a driver hits a high fatigue threshold, they will be sent a notification to take a break,” Porter explained.

At the same time, Grab has bigger ambitions, holding aspirations to use AI to help solve some of Southeast Asia’s “most complex problems”.

“We have one of the largest datasets in the region, which, combined with AI, can unlock new ways to change the lives of millions of Southeast Asians for the better,” Porter said. “One of the problems we’ve been actively working to solve is traffic congestion. Our driver-partners drive the most important roads in Southeast Asia multiple times a day.

“We’re working with governments and academic institutions like the National University of Singapore to develop new solutions based on this data to monitor and optimise traffic flow.”

Innovation engine

Porter is tasked with overseeing the development of agile platforms and machine learning, AI and data science capabilities, designed to help Grab deliver a “safer, more seamless and personalised” transport experience.

The technology executive leads and works with engineering teams across the company’s global network of six research and development (R&D) centres, spanning Singapore, Seattle, Beijing, Bangalore, Ho Chi Minh City and Jakarta.

“Our best innovations – the ones that get us the most excited – are the ones that make the biggest impact on the lives of the people we serve,” Porter said. “These innovations are not necessarily the most technically complex, and sometimes they hardly involve any tech at all.

“For example, we believe that every person, regardless of ability, has the right to financial independence. We have many driver and delivery partners that are hard of hearing or wheelchair-bound. We welcome them on our platform, and we’ve built various functionalities into the app to help them go about their jobs with ease.”

For partners with hearing loss, Porter said the app allows users to identify themselves as Deaf and turn off the call function – there is also an identification badge next to their name to prevent any misunderstandings with passengers.

Meanwhile for delivery partners that may be wheelchair-bound, Grab has helped install safety lights on wheelchairs, including rolling out options such as reducing the radius for the orders they get.

“Though many have told us through feedback that they don’t want to be treated differently,” Porter qualified.

Fundamentally, Porter said that while Grab is a business built on innovation, the organisation “doesn’t believe” in innovation for its own sake. Rather, projects are prioritised by the potential impact on “the lives of millions”, whether in “small or big ways”.

“At Grab every team is pushed to adopt a customer-first mindset,” Porter said. “Customer impact is probably our biggest metric for success, and many measures feed into this.”

For example, are customers getting their orders delivered within the given estimated time-frame? Did they get a ride on their first attempt at booking? Are driver-partners earning enough?

“We think holistically about our customers and what’s important to them,” Porter added. “In the technology team, as part of this effort, we keep our engineers as close as possible to our customers’ needs.

“As a CTO, I try to draw a direct line from what every engineer does to a customer benefit – if I can’t find one, we dig in. And in the digital age where time is of the essence, agile development is key to success.”

“Do it all at speed”

In looking back a decade ago, Porter accepted that it might have taken years before a feature improvement from an engineering team could reach the customer. Fast forward to 2019 and at Grab, this only takes weeks.

“It’s the speed of invention which is impacting all of the industries and categories we operate in,” he said. “The technology available today – AI, internet of things, cloud computing – gives us the opportunity to innovate faster than before to quickly provide benefits for our customers. To do that, we’re constantly experimenting.”

But Porter said running multiple experiments in tandem can be a “messy, complicated and expensive process”, with the business currently deploying hundreds of trials per week.

Hence the creation of the Grab experimentation platform (ExP), which Porter said provides “clean and simple” ways to identify opportunities, create prototypes, perform experiments, refine and launch products – “and to do it all at speed”.

Yet a time existed when implementing new technologies lacked such speed to market. Porter’s first introduction to the balance between features and prior investments was being one of 11 engineers on the Oracle database in the late 1980s.

“Oracle databases ran vast numbers of mission-critical databases that needed stability – but also needed innovative features and performance,” he recalled. “That’s when I learned that B2B customers care most about the product you’ve already given them being stable, secure, and iteratively improving – above and beyond new features, most of which would take quarters or years before going into production anyways.

“This was made even worse because we shipped software on tapes or CDs, and once shipped, it was incredibly difficult to patch.”

Today, the world is different, with Grab operating predominantly within the B2C space while adhering to a “customer-first” mindset.

“This means we can take a much more balanced approach to accounting for previous investments vs new features,” Porter added. “We regularly roll out new features, such as driver navigation or the ability to book rides and food and deliveries at the same time, first to a very small group of customers, allowing us to move fast and keep customer experience stable at the same time.

“Secondly, because we’re in the world of real-time updates, supported by the scalability and flexibility of cloud computing, we can roll out versions quickly with small increments of technology – which is much safer than the multi-year ‘Big Bang’ releases that were common until just a couple years ago.”