Meet Startup @TW

Taiwan’s SYSTEX Pairs with New York Startup to Improve Money Laundering Detection


Taiwanese information services firm SYSTEX Corp. says it will become the exclusive provider of anti-money laundering services for Graphen, an American startup that develops artificial intelligence using technology that mirrors the human brain.

This relationship announced at an August news conference will help SYSTEX position itself as the go-to place in Taiwan as the whole financial sector faces a third review in November for its defenses against money laundering. The 41-member Asia/Pacific Group on Money Laundering will do the review, which includes site visits.

Taiwanese financial companies are bracing for the evaluation, as low ratings will tarnish the domestic financial environment.

SYSTEX aims to use Graphen’s anti-money laundering service on helping clients stop money laundering by decreasing rates of false alarm, streamlining manual review of these alarms and identifying new channels that criminals might use. This service is also designed to lower risks of legal penalties and costs of regulatory compliance, according to an official news release.

SYSTEX Corporation

The 21-year-old company that also does mobile and cloud software services is now working largely on artificial intelligence, including an AI industry application exhibition earlier in 2018. Now, its AI technology is reaching the finance sector.

SYSTEX, which reported NT$16.2 billion in sales in 2016, supplies about one third of Taiwan’s anti-money laundering software systems, Senior Vice President Stony Fan said. Clients include major banks and securities firms. Most banks, Fan said, already have basic anti-money laundering systems but that they’re not enough.

“That is just the first step,” he said at the August news conference. “To further increase the systems’ effectiveness, banks still need AI to help pick out abnormal signals more precisely.”

Use of artificial intelligence in the detection of money laundering took SYSTEX to Graphen. The Taiwanese firm’s experience in serving major Taiwanese banks made it an ideal partner for Graphen in Taiwan. Graphen will work with SYSTEX on using AI to detect money laundering in Taiwan.

According to Graphen founder Lin Ching-yung, the startup was looking for a local company with accumulated data on financial organizations. Working with SYSTEX means they need to provide only the technology itself.

Graphen’s approach

Lin, an ex-IBM chief scientist, founded Graphen a year ago. It is headquartered in New York with branches and subsidiaries in five other cities including Taipei.

Graphen tries to stand out from other anti-money laundering system competitors by using graph-based AI platforms inspired by the human brain. The brain is seen as a graph of billions of nodes.

The current platform, which is used as a foundation of its AI trading service series, mimics all functions of the human brain, including but not limited to reasoning, understanding, and the ability to make decision.

“With a graph structure, computers can better emulate the full function of human brains, including memory, sensing, recognition, perception, comprehension, and strategy,” said Lin in a Meet Startup interview in June 2018.

Traditional databases require that information, such as the senders or the receivers in an anti-money laundering case, first be stored in separate tables and columns before a computer can analyze relations. This process has limits because of the amount of computing power required.

Graphen’s system makes it easier to detect links in a complicated network of money transactions including those between two corporations. These links can in turn help institutions root out any money laundering.

When working for one American client, Graphen’s AI anti-money laundering system picked out 50 possible money laundering cases that a basic system had overlooked, Graphen Asia GM Oscar Tseng said.

News source is from Business Next.