Independent Banker - July 2020 - 47

more about their own customer
by linking data among diff erent
products, " Pandit says.
The problem for many community
banks is that their data is often siloed
and in unnavigable formats. Usually,
it's generated as the bank's core
system processes loans, deposits and
other services, including debit card
systems, digital banking applications,
bill pay and much more. Some
community banks, depending on size,
could have hundreds of data sources.
So, how do banks access and
then connect all this information?
And, then, how can they be sure it's
not full of duplicate or superfl uous
data? Fintechs say that's where they
can help, with easy-to-navigate
dashboards, better-connected
products and more.
" The fi rst thing you hear from
[community bankers], is that, 'We
know we have a data problem ... yet, I
can't seem to get access to it, and once
I do, I don't know what to do with
it,' " says Keith Henkel, founder and
CEO of FI Works in Little Rock, Ark.,
a customer relationship management
(CRM) provider. " The thing that I
think more and more community
bankers are realizing is that [the data
problem] is a roadblock to doing a lot
of things. "
Once they've begun to take
advantage of their data, community
bankers may wonder how to make
best use of it. Artifi cial intelligence
(AI), or machine learning, is one
answer, and the technology is
at the core of many ThinkTECH
Accelerator 2.0 fi nalists' solutions.
Their software is increasingly relying
on the collective data of community
banks and other fi nancial institutions
to improve effi ciency and more. AI is
no longer science fi ction for banks of
any size.
AI-driven solutions
Take Fraud.net for example. The global
fraud detection company in New York
City mines bank data and continuously
feeds it into its security tools to catch
ever-evolving risk threats that a static
platform may not be able to detect.
" [AI and machine learning are]
behind chatbots, credit decisioning,
lending and, obviously, fraud
detection, to name a few, " says Kevin
Shine, vice president of sales and
partnerships at Fraud.net. " Fraud
detection is a data game. The more
data we can consume and decision
in real time, the better the outcomes
we can deliver to our clients and
their customers. "
In a similar boat is Finscend, which
off ers credit and debit card dispute
software that's built on a predictive
scoring engine. Card disputes can be a
huge cost and time suck for any bank,
says the fi ntech in Tel Aviv, Israel,
so its Banking Dispute Platform
relies on machine learning to take
in transaction data and real-time
information to assign an algorithmbased
confi dence score that the bank
can then use to auto decision the
dispute. Dispute resolution is a role
typically only a couple employees are
tasked with at a community bank, and
Finscend says relying on the engine
can make the process more costeff
ective, accurate and transparent.
" Our system takes away all that
subjectivity, " says Moshe Teren,
Finscend's chief technology offi cer
and cofounder. " Our software is built
from scratch; it's a tool that no one
has presented in the market today. "
On the fi ntech side, platforms built
on AI need data to be successful,
because the more information it
has to learn from, the better the
technology will perform. " We can only
build a platform on the data we've
been given, " says AK Patel, founder
and CEO of Lendsmart. " If you share
that with us, we can build a better
profi le from this existing borrower. "
In the case of Lendsmart, a loan
automation provider in New York City
specializing in mortgages, its solution
claims to reduce the mortgage
application process to 10 minutes. On
the bank side, AI allows Lendsmart
" We're not replacing the loan offi cer
or the underwriter; we're just making
your life easier. "
-AK Patel, Lendsmart
http://www.Fraud.net http://www.Fraud.net
Independent Banker - July 2020

Table of Contents for the Digital Edition of Independent Banker - July 2020
