Democratised AI expected to deepen inroads in Thailand

Economy December 09, 2017 01:00


4,053 Viewed

IMAGINE the likes of the Amazon Go shopping service with its cashier-less retail operation, Capital One’s voice-activated banking products or order fulfilment centres using robots and automation.

All these capabilities will be soon available more widely in Asia due to the democratisation of artificial intelligence (AI), according to Craig Stires, Amazon Web Service’s (AWS) chief of business development for Big Data and analytics. 

In Thailand, banks and retail operators have already prepared to take advantage of the new technology using machine learning models to get insights from the vast amount of data provided by customers and other partners. For example, The Mall group has joined hands with Siam Commercial Bank (SCB) to introduce a cashier-less supermarket service in Bangkok, while Central group has entered into a joint venture with, China’s No 2 e-commerce giant, to introduce a new retail experience. 

For the banking sector, Kasikorn Bank, SCB, Bangkok Bank, Bank of Ayudhaya, and Krung Thai Bank have applied with regulatory authorities to introduce QR code and related e-payment services to usher in the new era of a cashless society. 

Stires said at a media session of AWS’s re:Invent 2017 in Las Vegas last week that the democratisation of AI is made possible by the launch of new AWS products and current capability of real-time connections of IoT (Internet of Things) and mobile devices, cameras and the like for machine learning.

This will allow businesses such as banks and retail operators to provide a seamless experience to customers. 

Last week, AWS introduced five new machine learning services and the world’s first deep-learning video camera for application developers. One of its new services, the SageMaker, makes it easy to build, train, and deploy machine learning models, while its DeepLens camera gives developers hands on experience with machine learning. 

Other products include AWS Transcribe, Translate, Comprehend, and Rekognition Video which allow developers to build applications that turn speech into text, translate text between languages, as well as get insights from text and analyse videos for facial and other recognition. 

According to Stires, machine learning models of the AI technology work by turning a massive amount of data into models to create predictive analytics and, then, use data feedback over and over to create better models and better solutions. 

Cloud computing facilities can be used to train these machine learning models for seamless shopping or autonomous cars, Stires said, noting that voice recognition is also likely to be the next big thing for video and speech applications. 

Meanwhile, Thirapun Sanpakit, executive vice president of the Stock Exchange of Thailand, said SET is also using AWS’s cloud computing service, instead of its on-premise computers, for to boost scalability as a market boom leads a sharp rise of users via mobile applications.

 Most front-end services for stock price checking and graphs are now facilitated by cloud service while its back-end operation for processing buy and sell orders as well as portfolio record-keeping remain on the SET’s own computers. 

The SET earlier also launched the FundConnext platform – an online service for processing buy and sell orders for mutual funds to allow investors to access mutual funds more efficiently. 

In partnership with the mutual fund industry, the platform links brokers and sale agents of more than 12 mutual funds to serve retail and other investors better. 


The SET is also planning to launch another new service, called Live, which is an electronic trading platform for Thai tech start-ups wanting to mobilise funds from retail investors and institutional investors. 

The new service will use the blockchain technology for distributed ledgers, allowing electronic certification of start-up financial statements by auditing firms while using e-money for settlement. 

For its main board, the Thai stock market is in the process of introducing T-plus-2 for faster settlement on stock transactions within two trading days while smart contracts using the blockchain technology will likely be used in the near future.

Thirapun said the SET is becoming a more digital stock trading platform similar to Alibaba which serves as an e-commerce platform. 

According to Thirapun, blockchain is expected to be popular within the next five years, noting that Ripple, a cross-border money transfer platform, is a notable example of how blockchain can be used to provide new services.

In his opinion, AI and Big data are competitive tools for businesses as demand is no longer based on gut feeling, but is forecast by large amounts of data used to provide customer insights and to track customers. 

This will allow businesses to launch new products that can better meet or even exceed customer expectations. He said Thailand will require more data scientists and a new budgeting model for cloud computing services which are billed like electricity and other utilities.