Solution Brief
Financial Services Industry
Cognitive Analytics with Memory-Based Reasoning Systems
Addressing Financial Services
Challenges with Cognitive Analytics
From personalizing customer experiences to pinpointing undetected fraud to
identifying business efficiencies, Intel’s Saffron Natural Intelligence Platform*
helps financial institutions increase revenue and prevent loss
This solution brief describes how
to solve business challenges
through investment in innovative
technologies.
If you are responsible for…
• Business strategy:
You will better understand how a
cognitive analytics solution will
enable you to successfully meet
your business outcomes.
• Technology decisions:
You will learn how a cognitive
analytics solution works to
deliver IT and business value.
Executive Summary
Financial institutions are under increasing regulatory and compliance scrutiny,
competing in a growing atmosphere of disruptive technology and customer churn,
and searching for ways to drive efficiencies in their business operations. To solve
these problems, banks and insurance companies are turning to artificial intelligence
technologies to decrease the overhead of regulatory compliance, better understand
customer needs, and more effectively and efficiently manage the business.
Intel’s Saffron Natural Intelligence Platform* unleashes the next wave of artificial
intelligence with memory-based reasoning for cognitive analytics. Using associative
learning techniques, Saffron’s solution mimics our natural ability to learn, remember,
reason, and act. The Saffron Platform unifies heterogeneous data sets in an associative
memory store to learn and represent the knowledge inherent in the data. Coupled with
a memory-based reasoning engine, this approach provides a more accurate and adaptable
alternative to data modeling, especially when data sources are sparse and dynamic.
The Saffron Platform is architected end-to-end with Intel® technologies for highperformance cognitive analytics, customization toolkits, and ease of workflow integration.
By providing businesses with an analytics solution that learns and adapts
automatically, the Saffron Natural Intelligence Platform helps financial institutions
benefit from the knowledge in their data to win in a business environment that is
increasingly competitive, changing, and complex.
Saffron* Memory-Based Reasoning
LEARN
ACTION
ACT
REMEMBER
REASON
Figure 1. Intel’s Saffron Natural Intelligence Platform* delivers cognitive analytics
by replicating the human ability to learn, remember, reason, and act.
Solution Brief | Addressing Financial Services Challenges with Cognitive Analytics2
Solution Benefits
• Unification of diverse structured and unstructured
data sources
• Faster time to deep insights
• Greater accuracy
• Transparent actions and recommendations
Business Challenge:
Plenty of Data but Not Enough Insights
The financial services industry faces three primary challenges:
increasing governance, risk, and compliance (GRC) costs; an
increasing urgency to develop stronger customer relationships
and personalized services; and an ever-present need to increase
operational efficiency (Figure 2). In each case, there is no lack
of data—but legacy systems and processes are ill-equipped to
learn from the new forms and sources of data, transform the
data into knowledge, and extract rich insights with efficiency
and scale. Existing analytics tools and methods simply cannot
keep up with the increasing amount and complexity of challenges
facing banks and insurance companies.
Governance, Risk, and Compliance Costs
Financial services is one of the most heavily regulated industries,
with compliance costing some banks up to USD 4.5 billion
per year.1 GRC costs account for 15 to 20 percent of the total
“run the bank” cost base of most major banks.2 Much of this
GRC cost is associated with preventing fraud, such as money
laundering. An estimated USD 800 billion to USD 2 trillion is
laundered each year.3 Banks are highly motivated to curb fraud
and meet increasingly strict regulatory requirements—in
2014, banks paid USD 351 million in non-compliance fines
(about seven times the fines levied in the previous year).4
More than ever, banks are expected to “know their customer”
and provide the infrastructure, processes, and policies to
detect and prevent issues like money laundering. To increase
compliance and decrease fraud, banks must deploy advanced
analytics platforms that can handle both structured and
unstructured data, and quickly identify and respond to new
schemes that fraudsters are taking to hide anomalies.
voice, social media, email, browsing habits, and product and
service preferences can enable banks to increase customer
engagement. But this type of data mining requires cognitive
analytics to achieve the rich and individualized insights banks
and insurers need to deeply engage with their customers.
Business Efficiency
In the face of the rising cost of operations, competition from
innovators and start-ups, and customer demand for new
services, financial institutions must find ways to optimize
their processes, control their cost structure, and explore new
operating models. According to Deloitte research, banks that
have achieved continuous improvements in efficiency have also
generally experienced far greater gains in their share prices.5
Cognitive analytics can help discover inefficiency and drive
streamlined operations and discover new sources of value.
According to Deloitte research, banks that
have achieved continuous improvements in
efficiency have also generally experienced
far greater gains in their share prices.
Memory-Based Reasoning Can Solve
Financial Institutions’ Most Pressing Problems
The Saffron Natural Intelligence Platform* and its memorybased reasoning deliver a cognitive analytics solution that
addresses the business challenges facing financial institutions
today. Here are some examples of those challenges:
• Risk and compliance. Fast, comprehensive analysis of
heterogeneous data can lead to greater accuracy, lower false
positives, and prevent unforeseen fraud and money-laundering
actions. In one project, the Saffron Platform examined over
one-hundred thousand insurance claims and within ten
hours it uncovered a previously unknown fraud ring that
cost USD 2 million per year to the insurer.
Risk and
Compliance
Customer Engagement
Emerging players in the financial services market are using
disruptive technology to provide consumers with direct
access to services such as loans, insurance, and stock
brokering, cutting the bank out of the picture. The Internet,
mobile computing, and cloud-based services all make it very
easy for consumers to bypass traditional financial services
in favor of specialized offerings. To counter this trend, banks
and insurers must find a way to attract and retain customers
with new, personalized services that differentiate themselves
from “fintech” companies. Mining customer relationship
data, which can include a wide range of data types including
Customer
Engagement
Saffron
Natural
Intelligence
Platform*
Business
Operations
Figure 2. The financial services industry faces constant
pressures from regulatory scrutiny, customer churn, and
rising operating costs.
Solution Brief | Addressing Financial Services Challenges with Cognitive Analytics3
• Customer engagement. With a large and diverse customer
base and thousands of products and services, banks and
insurance companies are challenged to understand their
customers’ needs and respond in a timely manner. A
business can truly customize each interaction and more
deeply engage with their customers only by building an
associative memory store of all customer profiles and
preferences across the entire portfolio of products and
services. In one pilot project, the Saffron Platform improved
accuracy rates by 70 percent for product recommendation
across 160 different products and 8,000 distinct services.
All the while, the Platform accomplished these results in
10 weeks, compared to traditional modeling efforts that
took over 18 months and with less accuracy.
• IT operational efficiency and business operations.
The opportunities to apply memory-based reasoning to
IT operations are numerous. For example, Intel’s Saffron
Natural Intelligence Platform can be used to more quickly
find and resolve defects for software development, maintain
an expertise directory for sales staff, and provide talent
management for the human resources department. For
example, in one project, the Platform enabled a 10- to
15-percent increase in business operational efficiency by
accelerating software testing, with a 50-percent test
script deduplication.
Solution Value: Putting Data to Work
Financial institutions wanting to increase their competitive
capabilities in the market can turn to the Saffron Natural
Intelligence Platform. Analyzing data like the human brain
does, and with the ability to learn, remember, reason, and
act, Saffron’s knowledge store and memory-based reasoning
engine deliver distinct value to financial businesses:
Solution Architecture: Associative Memory
Store with a Memory-Based Reasoning Engine
The software solution stack for Intel’s Saffron Natural
Intelligence Platform is schema-free and can comprehend
disparate data types—it connects all the data points
regardless of source or structure. Source connectors and
extractors enable feature extraction using machine learning,
deep learning, and other artificial intelligence techniques
from Intel, third-party vendors, or the customer. Leveraging
Saffron’s patented associated learning algorithms, the
Platform creates an associative memory store, which can
then be queried using its powerful memory-based reasoning
engine and library of thought processes. Third-party data
visualization and other applications can be integrated into
the software stack for seamless decision analysis.
The software stack runs on a high-performance Intel®
architecture server platform, including the Intel® Xeon®
processor E5 and Intel Xeon processor E7 families, Intel® Solid
State Drives (Intel® SSDs), and Intel® Ethernet adapters (see
Figure 3). Other Intel technology integrates into the solution
stack to further optimize performance and security, including
the Trusted Analytics Platform (TAP) and the Intel® Scalable
System Framework (Intel® SSF).
The Platform can operate independently, integrate into an
existing on-premises infrastructure, or be delivered as a
cloud-based solution.
Data Visualization
Natural Intelligence Platform*
Thought Process Library
• Unify heterogeneous data sources (structured and
unstructured), leading to deeper insights and more
accurate predictions.
• Improve the productivity of financial crime investigators
by reducing false positives and generating results faster
than machine learning and deep learning solutions.
• Slash time to insight and resolution by learning instantly
and continuously, often reducing investigation times from
months to days and decreasing financial losses due to
fraud and compliance fines.
• Provide transparent, explainable, and auditable results
to move beyond black-box models and enhance the
organization’s decision-making confidence.
The Saffron Natural Intelligence Platform is deployable out
of the box—no data modeling or training is necessary. It can
be easily integrated with existing applications, processes,
and workflows to accelerate deployment.
Other Applications
Memory-Based Reasoning
Associative Memory Store
Techniques
Intel
Third Party
Customer
Source Connectors and Extractors
Structured Data Sources
Unstructured Data Sources
Intel® Architecture
Intel® Xeon®
Processors
Intel® SSDs
Intel® Ethernet
Adapters
Figure 3. Intel’s Saffron Natural Intelligence Platform* is
designed end-to-end and built on a high-performance
foundation of Intel® architecture.
Solution Brief | Addressing Financial Services Challenges with Cognitive Analytics
Conclusion
Artificial Intelligence technology is transforming the way
business is done and how enterprises create and deliver value
to their customers. The financial services industry has always
been at the forefront of adopting and implementing stateof-the-art technologies. Intel’s Saffron Natural Intelligence
Platform is the next wave of artificial intelligence, offering a
new approach to advanced analytics through memory-based
reasoning. Inspired by human intelligence, it transforms data
into knowledge, giving financial institutions the insight to
understand, the confidence to act, and the evidence to explain
the decisions used to tackle their biggest challenges.
4
Learn More
You may also find the following resources useful:
• Saffron Technology Website
• Intel® Artificial Intelligence
• Intel® Scalable System Framework
• Trusted Analytics Platform (TAP) Website
Find the solution that is right for your organization.
Contact your Intel representative or visit intel.com/FSI.
“Compliance Costs Community Banks $4.5 Billion Annually,” Drew Dahl, stlouisfed.org/on-the-economy/2015/december/compliance-costs-community-banks-billions
1
“You’ve Heard of Fintech, Get Ready for ‘Regtech’,” Matthias Memminger, Mike Baxter, and Edmund Lin, americanbanker.com/bankthink/youve-heard-of-fintech-get-ready-for-regtech-1091148-1.html
2
United Nations Office on Drugs and Crime, “Money-Laundering and Globalization,” unodc.org/unodc/en/money-laundering/globalization.html
3
“Regulators Issued Fewer AML Fines in 2014, But Packed a Bigger Punch,” Bao Nguyen, kaufmanrossin.com/news/regulators-issued-fewer-aml-fines-2014-packed-bigger-punch
4
5
“Improving Efficiency: The new high ground for banks,” nubank.com/stories/200907_ImprovingEfficiency/ImprovingEfficiency.pdf.
Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using
specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests
to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more information go to intel.com/performance.
All information provided here is subject to change without notice. Contact your Intel representative to obtain the latest Intel product specifications and roadmaps.
Cost reduction scenarios described are intended as examples of how a given Intel-based product, in the specified circumstances and configurations, may affect future costs and provide cost savings. Circumstances will vary. Intel does not guarantee any costs or cost reduction.
Intel technologies’ features and benefits depend on system configuration and may require enabled hardware, software, or service activation. Performance varies depending on system configuration.
No computer system can be absolutely secure. Check with your system manufacturer or retailer, or learn more at intel.com.
No license (express or implied, by estoppel or otherwise) to any intellectual property rights is granted by this document.
Copyright © 2017 Intel Corporation. All rights reserved. Intel, the Intel logo, and Xeon are trademarks of Intel Corporation in the U.S. and/or other countries.
Other names and brands may be claimed as the property of others.
*
0117/JWIL/KC/PDF
Please Recycle
335247-001US