Major U.K. Bank Has Signed on to Use Clinc’s Latest Fraud Tool

Clinc , a conversational AI company, has signed a deal with a major U.K. bank to use its new AI-based fraud tool, Bank Innovation has learned. Clinc could not disclose the bank's name, but Gualberto Camacho, director of product at the company, told Bank Innovation that it is "one of the three largest FIs in the U.K." This bank has made the tool available to its customers through Facebook Messenger and the bank's call centers, Camacho added. The new AI-based product is a fraud containment tool, built as a natural extension to Clinc's core virtual assistant product, Camacho said. <span style="display: inline !important; float: none; background-color: #ffffff; color: #333333; cursor: text; font-family: Georgia,'Times New Roman','Bitstream Charter',Times,serif; font-size: 16px; font-style: normal; font-variant: normal; font-weight: 400; letter-spacing: normal; orphans: 2; text-align: left; text-decoration: none; text-indent: 0px; text-transform: none; -webkit-text-stroke-width: 0px; white-space: normal; word-spacing: 0px;">By adding this capability, Clinc is helping its clients manage the volumes of fraud-related call/email-related queries, which betters customer experience but also helps the institution cut costs, he said. <span style="display: inline !important; float: none; background-color: #ffffff; color: #333333; cursor: text; font-family: Georgia,'Times New Roman','Bitstream Charter',Times,serif; font-size: 16px; font-style: normal; font-variant: normal; font-weight: 400; letter-spacing: normal; orphans: 2; text-align: left; text-decoration: none; text-indent: 0px; text-transform: none; -webkit-text-stroke-width: 0px; white-space: normal; word-spacing: 0px;">"The virtual assistant created for FIs was centered more around PFM capabilities," he said. "This fraud solution is adding to that to make it more comprehensive." As of last year, Clinc had 28 FIs worldwide that it was working with, Bank Innovation previously reported . In fact, Camacho was introduced to Clinc in 2016 when he was the innovation director at USAA Bank . The bank, of course, is one of Clinc's largest publicly disclosed partners. Clinc has powered USAA's capabilities on Amazon 's Alexa. USAA's mobile app has consistently topped numerous customer satisfaction lists for user experience. Capital One 's virtual assistant, Eno, is perhaps the only bank-built virtual assistant that is able to provide fraud containment and prevention solutions through text, mobile and call. Like USAA, Capital One, has consistently ranked high in terms of mobile customer service. Ken Dodelin, VP of conversational AI products at Capital One, told Bank Innovation that being able to detect and prevent fraud is an important and often overlooked component of customer experience. Dodelin declined to explain the precise technology behind Eno's fraud containment capabilities, except for saying that it is trained with AI and ML technology by using "tens of thousands" of customer transcripts. Aside from Capital One, fintechs like Clinc, Kasisto and Personetics , among others have carved a niche for themselves in working with FIs to build virtual assistants. But what separates Clinc from its competitors is its patented technology, Camacho said. "It's what allured me to join Clinc," he added. <span style="display: inline !important; float: none; background-color: #ffffff; color: #333333; cursor: text; font-family: Georgia,'Times New Roman','Bitstream Charter',Times,serif; font-size: 16px; font-style: normal; font-variant: normal; font-weight: 400; letter-spacing: normal; orphans: 2; text-align: left; text-decoration: none; text-indent: 0px; text-transform: none; -webkit-text-stroke-width: 0px; white-space: normal; word-spacing: 0px;">Clinc has a patented ability to extract context and interpret intent from natural speech, in any language. Unlike other conversational application makers that typically use parts of speech tagging, Clinc is more like Google Assistant in that it is not a rules-driven chatbot. It uses a technology called deep neural networks, which learns through data representations and not task-specific algorithms. "We train the AI model on what typically users might say," Camacho explained. "A lot of these platforms are trained to respond to keywords, we use a crowd vetting system where we reach out to people to collect everyday language and collect a wealth of utterances to train the platform." Further, the platform doesn't require thousands of examples to be trained, he said. Usually, anywhere between 300 to 400 "utterances" is sufficient to train the platform thanks to its <span style="display: inline !important; float: none; background-color: #ffffff; color: #333333; cursor: text; font-family: Georgia,'Times New Roman','Bitstream Charter',Times,serif; font-size: 16px; font-style: normal; font-variant: normal; font-weight: 400; letter-spacing: normal; orphans: 2; text-align: left; text-decoration: none; text-indent: 0px; text-transform: none; -webkit-text-stroke-width: 0px; white-space: normal; word-spacing: 0px;">end-to-end recurrent and bi-direction LSTM (long and short term memory) neural network architecture, which allows it to identify, remember and infer the user's intention and sentiment. This technology forms the basis of Clinc's latest fraud tool as well, Camacho explained. It is the "foundation for classifying the type of fraud," he said. <span style="display: inline !important; float: none; background-color: #ffffff; color: #333333; cursor: text; font-family: Georgia,'Times New Roman','Bitstream Charter',Times,serif; font-size: 16px; font-style: normal; font-variant: normal; font-weight: 400; letter-spacing: normal; orphans: 2; text-align: left; text-decoration: none; text-indent: 0px; text-transform: none; -webkit-text-stroke-width: 0px; white-space: normal; word-spacing: 0px;">For instance, a customer could say something like "I got this weird email from my bank," Camacho explained. In this case, the platform is able to understand: this is a fraud-related query, and it's taking place through email, he said. The actionable step is left to the discretion of Clinc's client, the bank. "So, however they want to deal with the next steps is up to them," he said. Clinc's U.K. bank client has been using this fraud solution since November, although Clinc only revealed this tool to the public two weeks ago. Since November, the bank was able to manage 90-95% fraud containment through its FB Messenger application, Camacho said. The product has been rolled out to 25 million of the bank's customers. Clinc plans on rolling the product out to other partners, too.

Major U.K. Bank Has Signed on to Use Clinc’s Latest Fraud Tool

Crypto Transactions are Up, Report Finds, but Institutions Remain on Sidelines

Hong Kong Permits Virtual Banks in Test to Traditional Lenders

Related Posts

MVB Bank bolsters vendor risk management via Kobalt Labs AI

Banking’s strong safeguards, regulations lower exposure to verification-request fraud

Nvidia introducing platform to police AI agents

Hong Kong Permits Virtual Banks in Test to Traditional Lenders

EMERGING FINTECH DIRECTORY

FinAi Podcast

SPONSORED

Build an Antifragile Strategy to Outperform the Market

How AI and Product Experts Turn Fuzzy Requirements Into Focused Dev-ready Roadmaps

Is Your Technology Supplier There for You?

Connect

Welcome Back!

Retrieve your password

Continue Reading with FinAi News Premium - Less than $2/Day

Major U.K. Bank Has Signed on to Use Clinc’s Latest Fraud Tool

Crypto Transactions are Up, Report Finds, but Institutions Remain on Sidelines

Hong Kong Permits Virtual Banks in Test to Traditional Lenders

Related Posts

FinAi Podcast

Connect

Welcome Back!

Retrieve your password

Unlock This Article

Continue Reading with FinAi News Premium - Less than $2/Day