AI seems to be well on its way to becoming the most overused buzzword of the tech industry, but don't be put off by the hype. Some fintech companies in Asia are actually making use of natural language processing or machine learning for detecting fraud and making investment decisions. I recently interviewed two CEOs—Simon Loong from the Hong Kong unicorn WeLab and Jianyu Tu from MioTech—to better understand some of the recent developments in AI in Asia's fintech industry.
Simon Loong in Welab headquarters in Hong Kong
Philippe Branche: First, could you describe your company in a few words?
Simon Loong: WeLab is a fintech company providing seamless digital financial services. Today, we have more than 800 employees and over 38 million registered users across Hong Kong, mainland China and Indonesia. Our key competitive advantages lie in our own proprietary risk management technology and advanced AI capabilities which help us effectively analyze unstructured mobile big data within seconds to provide innovative financial services and offer consumer financing solutions for customers in a matter of minutes. Before WeLab, to obtain credit, a consumer typically needed to jump through many hoops and provide a lot of paperwork to even be assessed, which reduced credit access. After which, the financial institution would take days, if not weeks to come to a credit decision. A good technology enabled credit model should be highly predictive, and this has enabled us to provide credit more efficiently and more effectively.
The company is highly profitable, with net profit reaching $17.7 million in 2017. And we expect to see higher profits in 2018 and 2019.
Jianyu Tu: MioTech is a provider of AI-driven Asian market risk intelligence and portfolio management solutions tailored to investors, delivering insights on financial risks in Asia. It is in this sector where developments in AI have a far-reaching impact: we solve problems by providing investors with a suite of analysis tools matched with comprehensive data coverage into companies, sectors, shareholders, supply chains, ESG and alternative datasets in Asia. Our clientele is 30% from mainland China, 70% from Hong Kong, and we are now expanding to Singapore.
What are the key recent fintech developments in your company?
Loong: We are proud to have obtained our virtual bank license in April 2019 and become the only homegrown fintech company in Hong Kong to establish a virtual bank. The virtual banking license application process lasted over 6 months, and WeLab eventually stood out from roughly 30 players who applied and became one of only eight licensees. The current prominence of fintech in Hong Kong is in a large part due to the foresight and supportive measures taken by the Hong Kong government and regulators. Hong Kong is really the leading city for finance and fintech in Asia. By applying innovative technologies including artificial intelligence and data analytics throughout customers’ financial journey, we aspire to revolutionize the banking experience for customers.
Tu: We caught sight of the trends in ESG–Environmental, Social and Governance–and impact investing early, which grants us access to a comprehensive China ESG database. Combined with our data, our software can perform industry and enterprise ESG risk analysis throughout an entire value chain to detect hidden vulnerabilities of related companies and individuals in real time. This AI-based analysis rests on publicly available sources, such as government websites, environmental protection sites, social media, and companies’ voluntary disclosures. We believe these new developments in ESG give us a relevant edge for our clients.
ESG Risk Metrics help investors benchmark companies, assess risk exposure and monitor risk trends ... [+]
How is AI relevant to your fintech company?
Loong: We are not only looking to achieve mere operational efficiency but also more intelligent risk management and better customer experience. Take fraud detection as an example, we utilize supervised machine learning to identify known fraud types, including those with highly complex patterns, and unsupervised machine learning to develop a labeling system for anomalies, which enables us to identify new fraud types. Our modeling approach includes both conventional linear regression and machine learning models. This helps us develop our credit scoring models, which are applied within our larger strategy of building multidimensional customer profiles. Within 5 months, we were able to develop our proprietary chatbot–“WeBot,” which utilized machine learning and natural language processing to provide predictive Q&A functionality. WeBot has the ability to enhance the customer experience whilst concurrently significantly reducing the dependency on human customer service representatives to handle customer enquiries.
Tu: We essentially use natural language processing (NLP) to abstract, synthesize information and translate relevant documents. Technology is now at a stage where it is intelligent enough to make sense of content in articles. For example, Bloomberg has 3,000 employees in India just plugging in data and manually reading articles. We only have 30 engineers based in Shanghai. My vision of AI is that it is not about algorithms anymore. Today, AI adaptation and applications are also key. We collaborate with universities such as HKUST or Tongji University in Shanghai to follow the latest AI trends.
The conversation has been edited and condensed for clarity.
