TheJakartaPost
Please Update your browser
Your browser is out of date, and may not be compatible with our website. A list of the most popular web browsers can be found below.
Just click on the icons to get to the download page.
Insight: How artificial intelligence helps firms stay relevant amid pandemic
Unfortunately, despite being around for quite some time, there are often misconceptions about what artificial intelligence can and cannot do.
Change Size
C
OVID-19 has put technology at the heart of many companies. Consumer behavior has shifted dramatically over the past few months, and ever more transactions are taking place online. While the pandemic has brought financial and operational challenges to all markets, technology, especially artificial intelligence (AI), proves that growth is still possible during times of crisis.
Unfortunately, despite being around for quite some time, there are often misconceptions about what AI can and cannot do.
Indonesia, with its large population and a deep smartphone penetration, presents a huge opportunity for data intensive technology. A combination of the massive amount of data and a significant increase in computing power is a perfect recipe for significantly accelerating the growth of AI applications across different domains.
Unifying and processing huge and diverse amounts of data is surely not an easy task, but this is the first hurdle companies need to clear in order to leapfrog to the next level of the innovation curve.
A 2018 study by McKinsey found that a late adopter of AI would risk losing 23 percent of cash flow by 2030. That is, if you do not start investing in AI today, most likely your competitors will. And the slower the adoption, the later you will be left behind.
At the core of AI lies machine learning, which allows machines to learn from a set of examples. Imagine that, instead of thinking about what rules to tell a machine, we can show examples for the machine to learn how to make scalable decisions. This is a great advantage, as it allows businesses to be more efficient and to scale their operations faster.
AI and machine learning, combined with the science of turning data into insightful information (aka data science), have become more important than ever in the “new normal” to guide innovation based on new market trends and consumer preferences.
To adapt to this ever-changing environment, businesses must know what their consumers currently need.
Gojek, for instance, leverages AI and machine learning to develop a GoFood food delivery service recommendation and search engine that can adapt to changing customer food preferences; and allocation engines that can figure out the best way to serve certain orders by taking into account the changing supply-demand curves.
And by leveraging these new technologies, one needs to build a comprehensive data team.
As the old saying goes, Rome wasn’t built in a day. Nor is your data team.
The next question would be: How can I build these capabilities? While there is no silver bullet to this, there are three key principles we need to follow to ensure our AI team consistently delivers a high-quality performance.
First, your early key hires in data would be very important. They will be the ones who set the tone for the years to come. Hiring people with experience in building data organizations would maximize the effectiveness of your team.
As your data team grows, it is important to ensure each member contributes meaningfully to the team. As in Gojek’s case, our data team comprises engineers, analysts, data scientists and decision scientists. Each contribute in their own unique ways and complement one another.
A good talent-search for your company’s growth is essential to set up a path to success.
Second, every data team needs to be agile. Given the nature of the work, the time to market for AI products need to be as short as possible. Your data team needs to work efficiently to reduce frictions to ship AI products and abstract away all the bookkeeping activities.
We are obsessed with automation. A lot of tools are developed in-house to increase the productivity and the speed of our executions. For instance, we internally have built a tool called Merlin to deploy machine learning models in a matter of minutes. This helps data scientists focus on model-building and reduce the bookkeeping that comes with model deployment, such as monitoring, logging, infra management, etc.
Third, hiring a team of agile data professionals is not the only recipe for high performance teams. Your data team needs to be empowered. We need to encourage all our employees to become scientists – that is, to always experiment with new ideas, fail fast and learn from it and, most importantly, make decentralized decisions based on data.
AI should be seen as a technology to augment human intelligence, helping us to make better decisions. Through data-driven/informed decision making, you can predict future trends, identify new opportunities, optimize your current efforts and produce actionable insight more efficiently.
It goes without saying that good collaboration between your data team and engineering as well as product teams is also needed to leverage on AI and machine learning.
The only constant in life is change. Nothing changes if nothing changes. The urgency is real, a big technological change is happening. Brace yourselves and be the first ones to be prepared for the changes.
----------
Gojek VP of data science. The views expressed are personal.
