Page 130 - Building Big Data Applications

128 Building Big Data Applications

infrastructure with the innovation of NoSQL databases and Hadoop, both of them
enriching the available layers with the bottom-most and top-most layers of infrastruc-
ture. In this new world, the implementation of AI-driven process engineering and ability
to execute as independent automated processes is where we bring in complex event
processing and operations algorithms which are very useful when implemented. Fig. 7.1
shows the entire thought encapsulated as a user experience.
Crowd engineering is a process engineering model which connects multiple unre-
lated processes, which are actually related and have financial impacts. Look at online
shopping for example, a consumer searches for products, finds what they are looking for
and either buy it or walk away from the entire process. The expectation is that one either
made a purchase or one did not, what if there were cookies that watched the entire
process, stole the credit card information for the consumer, and then transacted without
their knowledge to make purchases. The consumer is alerted by the bank that manages
the credit card and eventually is protected by FDIC rules for fraud. However, the banks
have to write-off the transaction using their fraud insurance. This issue needs to be
managed better and nobody should have to bear the consequences of fraud, but how can
we get to that end state? This is where the application of AI and machine learning will
help. The front-end browser and the associated financial pages can be encrypted with
algorithms that make it useless for a cookie or even an advanced persistent threat to steal
information. Have we done it? Yes or no, this is a later discussion.
AI and machine learning are having a major impact on banking, driven by vast
processing power and the continual development of new and more accessible tools and,
of course, the sheer volume of potentially useful, accurate data that all banks offer today.
Both retail and investment banks use machine learning in numerous contexts, from the
relatively mundane business of approving loans, to fund management, risk assessment,
and more. It is true that banks have used computer algorithms to trade stocks and shares
for decades. But that started when machines learning was more in a lab mode of
innovation. Today applications of machine learning in banking involve understanding
social media, news trends, and other data sourcesdbeyond stock prices and trades.
Today, an algorithm can play a part in calibrating a financial portfolio to the precise
goals and risk tolerance of the user, ensuring ideally, that a defined amount is earned by
a certain date from money invested. It can even autonomously adjust the management
of portfolios as market conditions change. An intelligent system scours millions of data
points, including granular trading information on companies around the world. It then
comes up with moneymaking strategies that it executes, as neural networks with min-
imal supervision.

•Learn
•Call
Backend
Center
Business •Purchase •Detect Mobile
AI
Event App •IVR
•Fraud •Alert
Process
•Alert
•Predict
FIGURE 7.1 Crowd engineering.