from emails and recorded customer lines
which would indicate that customers didn’t
know or understand which accounts they had
with a bank. For example, semantic analysis
performed on calls, emails, and other employee-to-customer interaction, will identify consistent
themes within complaints and conversations. If
customers complain that they are being charged
for accounts they did not open, or they are finding accounts opened in their name, this would
determine a more widespread issue.
■ ■ ■ Employee termination for fraud or ethical
violations. While this isn’t necessarily a “data
science” technique, correlation analysis can
identify a problem with front line employees.
For example, if management saw employees
being fired for ethical violations in numbers
that are significantly higher than the past,
or higher than industry averages, this action
should be a warning flag that something inappropriate is underway.
The tried and true data science techniques of
finding anomalous patterns of behavior would
be perfectly applicable to a case of employee
fraud. Organizations need to be proactive and
play the offense. While data analytics helps
organizations know what has happened and
respond, it can also thwart malfeasance. Send-
ing an automatic email to customers when an
account has been opened tends to be standard
practice in most industries nowadays, how-
ever, some banks are still behind the times. If
a customer receives an email about an account
that has been opened, that customer can im-
mediately respond and the situation is promptly
resolved. It’s a small fix, but when implemented,
it can ensure that accounts are not opened
fraudulently or in error, as employees are more
likely to be more scrupulous where there are
checks and balances.
Banks should also use data science to know
what their baseline—their normal—looks like.
Banks should do analysis and consistently review characteristics for their account holders to
learn where there are anomalies. How diverse
are email addresses for single account holders
with multiple products? How many active and
inactive accounts does the bank typically have?
Where are accounts typically opened, and how
many are typically opened and closed? What IP
addresses are being used for banking and where
are the IPs? Are the same IPs being used to access different accounts? Knowing what business
your bank has, beyond just revenues, will alert
you to deviations and in turn, executive management can take action in a timely manner.
Considering how the technologies of data
science are well-suited to identify unethical
employee activity, the question will become
management’s will to find and put a stop to it,
and the Board’s independence to demand the
appropriate governance controls. With an effec-
tive data science-powered solution in place and
by asking the right questions, banks can be able
to identify and prevent internal employee fraud
and avoid serious damage to its brand identity,
as well as millions of dollars in fines. ■
ABOUT THE AUTHOR
DAVID MCLAUGHLIN is CEO of
QuantaVerse, a data science company
founded specifically to help the
financial services industry.
QuantaVerse, Inc. is the emerging
leader in data science-powered risk reduction and
revenue growth solutions, purpose-built for the
global banking industry. Founded by financial
services industry veterans and innovators,
QuantaVerse solutions employ proprietary data
science algorithms, integrate and filter internal
bank data and related external data–including
public Internet data, unindexed deep web data and
government and commercial datasets–to help the
global banking industry to significantly improve
their compliance with AML, KYC and BSA
regulations and requirements. David can be
reached at dmclaughlin@quantaverse.net, or for
more information, visit www.Quantaverse.net.
Data science, including
big data, artificial
intelligence and machine
learning, can enable
institutions of all
industries and sizes to
better manage risks, to
find fraud and criminality,
and to therefore support
the needs of compliance,
management and
governing boards.
D
h
p
r
a
e
a
KU
RO
KS
TA
/
SHUTT
E
RS
TO
CK
