Ahead of the Big Data & AI Leaders Summit Singapore 2019, we interviewed one of the keynote speakers - Angus Kong, AVP, Data Science at Tokopedia. Angus manages the data science department in multiple countries. He oversees the Artificial Intelligence and Data Science development and strategies for Tokopedia including in-house technologies, industry, government and academy partnerships.
How has data gathering and analysing changed since you first joined Tokopedia?
I have just joined Tokopedia for less than half a year. My time with Tokopedia is not yet long but the change is already large. Tokopedia has been putting much effort internally to make all the data storage and access in a standardised way. We are building multiple data platforms for different types of data in Tokopedia. Now the data platforms are already covering more than 90% of our data regarding products, users and merchants. When I joined the company, only a few of them were in the planning stage.
Is your data team centralised or decentralised and why?
Our data team is centralised. This is still the most efficient way for us to work with data. With all the data teams centralised, we can easily share our knowledge and understand all the plans and changes in this mobile area. It's also our data team that is still growing rapidly and having a centralised team can help us manage the resource and stay flexible. I believe we will change the way we organise data teams in the future. I don't think there is a perfect answer here.
What are the eCommerce technology trends that the industry should look out for?
Fintech has been popular in recent years. With more data about the users, we can understand more the financial profile of them. I believe soon Fintech will be part of eCommerce technology. In fact, many people already saw the natural integration of Fintech and eCommerce. It also improves the user's eCommerce experience a lot with Fintech.
Where do you see AI or machine learning leading your industry in the future?
Tokopedia is determined to invest in AI or machine learning technology. As our mission has always been "democratise commerce through technology", we will enable everyone in our ecosystem to work with the AI technology from us and build the technology as a bridge for everyone to achieve. AI technology will be a commodity on our platform.
What are the main challenges to your role this year?
Growing the team, looking for great talents is always difficult. This year we opened our data science office in Singapore and I have been working on growing the team in Singapore, this subsidiary office is a part of our RnD, to support Tokopedia business in Indonesia. In Tokopedia, it's always challenging to grow a team. We value our culture very much and always put it in the first place when we interview people.
What will you be discussing in your presentation?
Many companies are building their own software products. The process of software development is a hot topic. People always want to know how to improve the process and try to be more efficient. I'll talk about how to build AI software products. From a software development perspective, to build AI models in a product, things are definitely more challenging and unpredictable along the development process. It is very interesting to see the mistakes we made and learn from them.