The fast advancement of technology has made AI technology, big data, and machine learning
more prevalent. These, combined with the impact brought about by Covid-19, have accelerated
the migration of customers to digital channels. Under such a trend, E.SUN, with AI as the
key technology, provides customers with real-time, safer, and more convenient digital
financial services through digital transformation, which was made possible by our culture of
experimentation and cross-functional teams. To fully utilize the capacity of AI technology,
E.SUN built
the first Machine Learning as a Service platform in the financial industry as a bridge that
connects AI technology and business procedures. It provides a variety of innovative AI
solutions, including consumer protection, fund target recommendations, internal control
detection, and abnormality alert among others.
E.SUN Wallet
Following the launch of the "Instant Card-linking" and "Card Management" functions in
2019, in 2020, three major functions including "Card Management 2.0," "Consumption
Category Inquiry" and "Scan & Go" were further made available on E.SUN Wallet,
continuing to provide convenient mobile payment services for creating a seamless
card-using experience for customers.
(1) Card Management 2.0
The "Card Management" service that extremely popular with the digital community was
enhanced by adding the "single designated limit setting" function. Customers can set
different purchase limits according to the usage habits for each card, providing a more
convenient card management experience. For example, customers can set a single limit of
NT$1,000 for credit cards used for purchases of small amounts. This allows the system to
help control spending limits, and reduces the risk of card loss and theft.
(2) Consumption category inquiry
Through data analysis and user feedback, E.SUN understands that customers like to check
consumption status in real time. In response, the bank launched a new function entitled
"Consumption Category Inquiry" which divides consumption into general consumption,
online general consumption, overseas general consumption, general F&B consumption,
supermarket QR code payment and other categories. This allows customers
to easily track their spending through the classification function and calculate the
reward points for various types of consumption.
(3) Scan & Go
The payment QR code combines carrier and membership functions. When customers use the
"E.SUN Wallet" to spend at cooperating convenience stores, they only need to scan the
payment QR code once to complete multiple tasks such as payment, invoice storage,
points accumulation, and obtaining member discounts, thereby enjoying full access to the
conveniences of mobile payment.
E.SUN Extremely Fast Loan
In 2015, E.SUN became the first financial institution
to launch e-Click loans, providing customers with online loan application services,
saving customers the hassle of visiting a branch and saving their valuable time while
contributing to reducing carbon emissions with paperless operations. In February 2020,
E.SUN again engaged in innovation and pioneered a fully digital process, "E.SUN
Extremely Fast Loan" using AI technology to create a loan process with zero manual
intervention through which customers can obtain funds in as little as 58 seconds. In
addition, different from traditional credit limits entailing amounts of at least
hundreds of thousands, Extremely Fast Loan enables loan applications of as little as
NT$20,000 and can also be flexibly repaid, providing customers with both speedy and
flexible loan services. As of the end of December 2020, the E.SUN Extremely Fast Loan
disbursement amount has reached more than NT$30 million, and is able to save nearly 95%
of case costs compared to loans made using paper-based processes.
AI Selection of Fund
In 2015, E.SUN became the first financial institution
to launch e-Click loans, providing customers with online loan application services,
saving customers the hassle of visiting a branch and saving their valuable time while
contributing to reducing carbon emissions with paperless operations. In February 2020,
E.SUN again engaged in innovation and pioneered a fully digital process, "E.SUN
Extremely Fast Loan" using AI technology to create a loan process with zero manual
intervention through which customers can obtain funds in as little as 58 seconds. In
addition, different from traditional credit limits entailing amounts of at least
hundreds of thousands, Extremely Fast Loan enables loan applications of as little as
NT$20,000 and can also be flexibly repaid, providing customers with both speedy and
flexible loan services. As of the end of December 2020, the E.SUN Extremely Fast Loan
disbursement amount has reached more than NT$30 million, and is able to save nearly 95%
of case costs compared to loans made using paper-based processes.
Watch-listed account detection system
To counteract the endless stream of fraud cases, traditional financial institutions have
been using manual flagging of watch-listed accounts to prevent further fraud. However,
due to low accuracy, additional manual phone calls to customers are required to
eliminate misjudgment of account transactions. The lengthy confirmation process causes
customer distress and a time gap in protection.
By introducing the MLaaS system, E.SUN built a model using machine learning for
automatically identifying watch-listed accounts with accuracy 40 times higher
than that of traditional manual flagging. This not only reduces the frequency of
interrupting customers with manual phone calls, the model judgment result can also be
directly connected to the accounting system to speed up flagging of the watch-listed
account, thereby achieving the effect of real-time synchronization of information, and
enhancing the timeliness and comprehensiveness of fraud prevention.
Credit card fraudulent transaction detection system
Through the use of AI technology, the bank is able to overcome sporadic and complex
methods of fraud that are difficult to prevent. E.SUN generated a model database of
nearly 400 factors through exploratory data analysis, and used the Gradient Boosting
Decision Tree (GBDT) algorithm to improve the generalization and predictive ability of
the model. Data cache technology was then further applied to perform abnormality
monitoring of the latest transaction data at all times so that each credit card
transaction can undergo risk evaluation within 0.1 seconds, thus accurately and quickly
protecting the security of customer transactions.
Mobile PayPass
Due to the impacts of the COVID-19 pandemic, the volume of online and mobile transfer
transactions grew significantly in 2020; in particular, the number of mobile transfer
transactions grew by as much as 65%.
In response to changes in consumer habits, E.SUN launched the E.SUN PayPass API. As long
as online shoppers can interface with E.SUN PayPass, they can directly call the mobile
banking apps of multiple financial institutions on the online shopping platform, saving
the time needed to integrate the mobile banking APPs of various banks. Additionally,
consumers do not need to enter the transfer account number and amount on their own;
payment can be completed as long as customers pass the fingerprint or password security
authentication of mobile banking, thereby offering customers a more efficient shopping
experience.
