UK-based Featurespace has won a major contract with HSBC, which has started deploying the company’s machine learning platform to help identify financial crime.
The contract will see the Cambridge-based firm’s Aric platform used to combat money laundering, automatically detect anomalies in client behaviour and prioritise these for staff, who can receive up to one million fraud alerts per month; an overwhelming deluge of emails that, ignored, can result in major compliance issues.
HSBC joins ClearBank, TSYS and Worldpay in using the platform, which claims to reduce false positives by 76 percent. The company also offers a white label version of Aric for users to customise and add their own branding too.
Read this: Payments Sector Wins Open Banking Reprieve with 18 Month SCA Extension
Martina King, CEO at Featurespace said: “In the fight against the ever-evolving threat of financial crime, collaboration is key. These firms’ application of our technology will help them in optimising the detection of suspicious activity, ultimately protecting their customers from the global impact of financial crime.”
It recently added predefined PSD2-specific data feeds and configurations to Aric – including rules and adaptive behavioral analytics that automatically decide which transactions to escalate to Strong Customer Authentication.
Featurespace: Just 1% of Alerts Red Flagged for Human Investigation
Featurespace said its analysis estimates that less than one percent of the estimated million monthly emails (for a large bank) are ‘red flagged’ for human investigation – around 10,000 a month; each individual alert can then take a week to investigate.
The outcome: a typical large bank requires 400,000 staff hours a month to investigate all its flagged financial crime alerts. Featurespace is among the companies using machine learning to winnow down and prioritise the alerts, saying it can help banks spot potential criminal activity a month earlier than rules-based systems.
HSBC Ramps Up Tech Investment
The bank has been making significant investments in machine learning, both on the compliance side and to help it develop new client products.
In June it signed up Canada’s Element AI to help it deliver insights from a newly collated data lake comprising 10 petabytes of corporate and institutional client data from 1.6 million clients.
Last month (August 2019) meanwhile, the bank landed the CloudBees CEO Award for DevOps transformation excellence.
CloudBees CEO Sacha Labourey noted at the time: “They evolved from being not only a financial services organization, but to truly being a leading IT organization, too. In an organization the size of HSBC, it’s not easy to implement a DevOps transformation. But they are doing it. They are determined, gritty, tenacious – and committed.”
Read this: Introducing The Intelligence Network: BAE Systems’ 1,500-Strong Coalition to Tackle Cyber Fraud
