Data Analytics: Implementation PDF Free Download

Name: Data Analytics: Implementation PDF
Author: Michael Williams

1 / 56

2 views•56 pages

Data Analytics: Implementation PDF Free Download

Data Analytics: Implementation PDF free Download. Think more deeply and widely.

1Insight. Oversight. Foresight. SM

Michigan Texas Florida  North Carolina

Data Analytics:

Implementation

Presented by:

Robin D. Hoag, CPA, CGMA, CMC

Shareholder, Financial Institutions Group

Region 3 Meeting

September 18 –20, 2019

Lansing, Michigan

Financial Institutions

Group

Today’s Objectives

•Introduction to data analytics

•Why it’s important to Internal Auditors

•Overview of the key elements, attributes, challenges

•Steps in the data analytics process

•Data analytic tools

•Roles and responsibilities

•Applications for Internal Audit

•Resources

Financial Institutions

Group

What is Data Analytics? Definition

•The process of inspecting, cleansing, transforming,

and modeling data with the objective of highlighting

meaningful information, suggesting conclusions, and

supporting decision-making.

Financial Institutions

Group

Data Analytics: More!

•Problem-solving process

•Extracts insights

•Historical, real-time, or predictive

•Data analytics (DA) can be:

•Risk-focused (i.e., controls

effectiveness, fraud, waste,

policy/regulatory non-

compliance)

•Performance-focused (i.e.,

increased sales, decreased

costs, improved profitability.)

Financial Institutions

Group

FOCUS on Relationships

•Identify and interpret relationships among variables to

facilitate decision-making using the Five W’s:

•Who

•What

•Why

•Where

•When

Financial Institutions

Group

Why Is Data Analytics Important To

Internal Audit?

Strategic Area Enhancement

Credit Union Expectations Audit coverage, quality, business

impact, on a finite audit budget

Regulatory Expectations Stronger assurance and quantifiable

results

Competitive Landscape Strengthen capabilities

Seek new talent

Increased Value Deeper discussion on issues

Develop/strengthen relationships

Talent Development Strengthen business skills

Appeal to other staff members

Boost recruiting

Financial Institutions

Group

Why Is Data Analytics Important To

Internal Audit?

•Internal audit departments leverage data analytics in

order to:

•Identify additional risk

•Increase scope and coverage -assurance

•Better understand existing risk through data

•Identify holes in control systems

•Augment limited resources

•Provide insights to management

Financial Institutions

Group

Why Is Data Analytics Important To

Internal Audit?

•Some areas that benefit from data analytics:

•Loan operations

•Accounting and general ledger transactions

•Accounts payable

•Payroll

•Deposit transactions

•Compliance

•Legal and regulatory compliance

9Insight. Oversight. Foresight. SM

Challenges To Using Data Analytics

Financial Institutions

Group

People

•Limited resources (financial and human) to execute on

a sustained basis

•Appetite for investment in time and training needed to

develop an effective DA / AI process

•Someone needs to create, run, and maintain queries

•Proficiency using analytic software

•Proficiency in performing analysis

The top barrier for implementation of big data analytics is

“inadequate staffing or skills for big data analytics.”

(Source: The Data Warehousing Institute (TDWI))

Financial Institutions

Group

People: Senior Analysts

12 Insight. Oversight. Foresight. SM

Past, Present, and

Not So Distant Future

Financial Institutions

Group

“Computers will disrupt work habits

and replace old jobs with ones that are

radically different.”

Isaac Asimov predicts 2019 (in 1983)

Financial Institutions

Group

We are in the Midst of a Data-Enabled Shift that is

Transforming Our Stakeholders and How We

Serve Them

We have transformed from a “Digital Driven Economy”

to where the “Economy is Digital”

Financial Institutions

Group

Impact of Analytics

Financial Institutions

Group

Artificial Intelligence: Introducing the

Technology

“I think the simplest definition of AI is that it’s a machine

or a computer exhibiting the characteristics that we

would normally associate with human intelligence”

Robin Grosset, CTO

MindBridge AI

Financial Institutions

Group

Let’s Get a Bit Technical

Artificial Intelligence

1956+

Early

artificial

intelligence

stirs

excitement

Machine Learning

1980+

Machine

learning begins

to flourish

Deep Learning

2010+

Deep learning

breakthroughs drive AI

boom (Google Cat)

Financial Institutions

Group

Comparing Ensemble AI with

Traditional Rules

Rules alone:

Ranked risk score

Machine learning:

Ranked risk score

Recent real-world example:

In 1.6M transactions, flagging those of interest

•Machine learning identified the fraud as 34th riskiest transaction

•Traditional rules ranked the fraud ~31,000th in priority to examine

900X improvement

Financial Institutions

Group

What Can Machines Do Better Than

Humans?

Financial Institutions

Group

Trust Within AI-Enabled Analytics

Financial Institutions

Group

Traditional CAATs

Traditional CAAT tools are rules-based, with limited coverage:

Financial Institutions

Group

Using AI

Financial Institutions

Group

The Art of Evaluating AI

Its potential for good and progress will only fully emerge when algorithms

become explainable in simple language

Financial Institutions

Group

Ta k e -Aways

•The pace of technological disruption is accelerating and will

continue to impact us

•Big data and improved processing power, drive the

development of AI applications

•Big data is underutilized in the internal audit space and AI can

help harvest those opportunities

•AI can help with narrow tasks (data processing, risk scoring,

search) with the human focusing on cognitive tasks

(communicating and advising)

•Adopting AI is a journey and not a silver bullet. Educate

yourself and start small.

25 Insight. Oversight. Foresight. SM

Data Analytics Tools

Financial Institutions

Group

Data Analytics Tools

•The right data analytic software will:

•Handle large data sets efficiently

•Integrate well with big-data

•Include wide-array of analytical and statistical functions and procedures

•Be relatively easy to program

•Log procedures performed on data

•Allow users to easily re-run analysis with minor changes

•Be scalable with regards to the platform

•Ensure the vendor’s vision is inline with the organization’s vision

•Include training and support

Financial Institutions

Group

Data Analytics Tools

•Basic

•Microsoft Excel

•Microsoft Access

•Integrated query tools

•PeopleSoft

•SAP

•Oracle

•JDE

•Specialized DA visualization

software

•Tableau

•Qlikview/Qlik Sense

•Data analytics

•Mindbridge

•Treasure data

•NICE

•Periscope

•Zoho

•DXC

•Sisense Data analytics

•Inflow

•AWS

•Specialized auditing software

•ACE, IDEA, Arbytus, SAS

28 Insight. Oversight. Foresight. SM

Data Analytic Team’s

Roles and Responsibilities

Financial Institutions

Group

Key Roles and Responsibilities:

Internal Audit

Splitting the analytics roles -essential ingredients…

1. Audit management & staff

Provides comprehensive understanding of the audit objectives

Identifies opportunities to introduce data analytics into the audit

process

Drives demand through personal insights and relationships

Keeps focus on solving audit related issues

Financial Institutions

Group

Key Roles and Responsibilities:

Internal Audit

2. Data analytics SME

Proficient in use of DA tools and is able to design queries and

manipulate data easily

Experienced auditor with a knack for analysis

May have knowledge of advanced statistical topics and modeling

Excellent problem solving skills

Financial Institutions

Group

Key Roles and Responsibilities:

Internal Audit

3. Data specialist

Strong programming and coding proficiency

Has been a database administrator or systems analyst

Has spent time as developer and has built applications

Expertise in core IT related functions in querying, data extraction,

cleansing, and manipulation

32 Insight. Oversight. Foresight. SM

Data Analytic Applications for

Internal Audit

33 Insight. Oversight. Foresight. SM

Practical Examples

Financial Institutions

Group

Data Analytics Applied to Accounts

Payable

•AP tests can be designed to address risks, cost

savings and/or recoveries

•Data analytic tests can be designed to identify any of

the following:

•Improper disbursements

•Duplicate payments

•Unapproved purchases

•Payments for items not received

•Payments in excess of approval levels

•Missed discounts or credits

Financial Institutions

Group

AP Analysis: A Few Ideas

•Improper payments or questionable disbursements

•Detect duplicate payments using dates, payees, vendor invoice numbers and

amounts*

•Identify invoices or payments to vendors without a valid purchase order*

•Look for invoices from vendors not in approved vendor file

•Find invoices for more than one purchase order authorization*

•Identify multiple invoices with the same item description*

•Extract vendors with duplicate invoice numbers*

•Look for multiple invoices for the same amount on the same date*

•Find invoice payments issued on non-business days (Saturdays and Sundays)

•Identify multiple invoices at or just under approval cut-off levels

•Identify credits issued by or outstanding with vendors

•Identify goods invoiced and paid, but not shown as being received

•Look for payments to vendors not on contract.

* signifies potential for recoveries

Financial Institutions

Group

AP Analysis: A Few More Ideas

•Look for multiple payments to the same vendor on the same

date or for the same amount (excluding recurring charges, such

as rent)*

•Stratify vendor balances, check amounts, invoice amounts, PO

amounts, etc., for unusual trends or exceptions*

•Calculate and validate annualized unit price changes in

PO/payments for the same product over time*

•Review sequence of check numbers for gaps

•Identify payments where no discount was taken*

•Review changes to the vendor master file

* signifies potential for recoveries

Financial Institutions

Group

AP Analysis: A Few More Ideas

•Look for multiple payments to the same vendor on the same

date or for the same amount (excluding recurring charges, such

as rent)*

•Stratify vendor balances, check amounts, invoice amounts, PO

amounts, etc., for unusual trends or exceptions*

•Calculate and validate annualized unit price changes in

PO/payments for the same product over time*

•Review sequence of check numbers for gaps

•Identify payments where no discount was taken*

•Review changes to the vendor master file

* signifies potential for recoveries

Financial Institutions

Group

Accounts Payable Schemes

•Phantom vendors

•Match names, addresses, phone numbers, bank accounts

and taxpayer identification numbers between vendor source

documents.

•Verify existence of vendors using a PO Box for an address

Financial Institutions

Group

Accounts Payable Schemes

•Kickback or conflict-of-interest

•Vendor prices greater than standard

•Price increases greater than acceptable percentages

•Continued purchases in spite of high rates of returns, rejects,

or credits

•High volume purchases from one vendor

•Frequent change orders

•Identify payments to vendors with same names, addresses,

phone numbers, etc., as employees

Financial Institutions

Group

Accounts Payable Schemes

•Bidding and contracting

•Patterns of rotation among vendors

•Bids that are exceptionally lower than those of other vendors

•Low winning bids followed by numerous change orders

•Excessive use of one contractor in a competitive field

•Patterns in awards to vendors

•Identical bids

•Multilateral drops in bid prices (accompanied by the entry of new

competitor)

•Competitors with the same addresses, principals, sales agents, phone

numbers, etc.

•Vendors with same names, addresses, phone numbers, etc., as

employees.

Financial Institutions

Group

Other Applications for Data Analytics

•Accounts Receivable

•Valid sales orders

•Accurate product pricing

•Authorized shipments

•Proper invoicing

•Valid cash receipts

•Timely collections & write-offs

•Sales contract compliance

•Other adjustments

•Payroll

•Accurate & authorized payments

•Timely & accurate hires &

terminations

•Reasonable OT & commissions

•Proper timekeeping & attendance

•Search for non-existent

employees and other payroll

schemes

•Comparison of periods for

unusual trends

Financial Institutions

Group

Other Applications for Data Analytics

•General ledger

•Journal entries

•Closing activities

•Adjustments

•Master files

•Members

•Loans

•Travel and entertainment

•Purchasing cards

•Data quality

•Reasonable

•Within expected range

•Validity

•Complete

•Compliance

•BSA/AML

•Money service business

Financial Institutions

Group

Rules Based View of Audit Procedures

(Legacy C AATs )

A legacy but still common practice

encourages the use of audit testing tools

where each test is done one-by-one.

Each test is performed in isolation and

then examined by the auditor to look for

issues.

This focuses the auditor on specific

issues and helps to verify the presence

of controls and good accounting

practices.

These techniques increase the odds of

finding anomalies. But this can be

improved

Financial Institutions

Group

Rules Based Tests

Typical audit test pattern:

Tests performed one by one and human auditor inspects results

     

Financial Institutions

Group

Human Hypothesis Generation

What can I say about these transactions? What else is

interesting?

A Human

Centric View of

Analysis

Financial Institutions

Group

Rules Based Testing: Combinations

By considering all tests together and viewing a

combined result score, you can see higher risks

floating to the top which fail more tests.

More Risk Factors Fewer Risk Factors

Ranked Interest Score

Conclusion: Combining the test outcomes and understanding the intersect

produces better results when trying to understand transaction risk.

Financial Institutions

Group

A Real World Scenario: Rules Combined

Rules-based testing can create wide (large) buckets

of transactions

Financial Institutions

Group

A Real World Scenario: Machine Learning

Machine learning techniques create much better differentiation

of transaction risk factors.

Ranked Risk Factor Scores Histogram of Risk Factor Scores

Good distribution and separation enables

appropriate focus and use of time

Smooth range of risk

scores, no large buckets

Financial Institutions

Group

Quick Comparison:

Machine Learning vs. Rules

Conclusion: Machine learning is proving a better tool to understand risk

factors and it produces better outcomes than rules based approaches alone.

In an example scenario “Unusual Cash Disbursements”:

•Rules: flagged a transaction as normal putting it in the 30th percentile of risk

•Machine Learning: flagged the same transaction in the 3rd percentile of risk

Financial Institutions

Group

Machine Learning: Rare Flows

Characteristics of potentially inappropriate journal

entries

•Made to unrelated, unusual (e.g., unusual combinations of

debits and credits), or seldom-used accounts

Audit Rationale

Financial Institutions

Group

Machine Learning: Rare Flows

Algorithm Finds: Unusual Journal Entries based on frequency of account interactions

For each and every transaction (100%), measure how often they occur and flag rare interactions

Algorithm Class: Unsupervised Machine Learning

Questions?

Financial Institutions

Group

References

ACFE (2018). Report to the Nations. Retrieved 4/30/2019: https://www.acfe.com/report-to-the-

nations/2018/

AICPA (2018). Beyond robotics: How AI can help improve audit process. Retrieved:

https://blog.aicpa.org/2018/08/beyond-robotics-how-ai-can-help-improve-the-audit-

process.html#sthash.3Lzmii1i.uAOqhSbF.dpbs

AICPA (2019). A CPA’s Introduction to AI: From Algorithms to Deep Learning, What you Need to

Know. Retrieved:

ARRIA (2018). Why the Finance Industry is Ripe for AI Disruption. Retrieved:

http://blog.arria.com/why-the-finance-industry-is-ripe-for-ai-disruption

https://www.aicpa.org/content/dam/aicpa/interestareas/frc/assuranceadvisoryservices/downloadabl

edocuments/cpas-introduction-to-ai-from-algorithms.pdf

BBC (2018). Carillon: The audit industry’s existential question. Retrieved:

https://www.bbc.com/news/business-44201251

Change.org(2019). Change Oxford English Dictionary’s Archaic Definition Of The Word

‘Accountant'. Retrieved: https://www.change.org/p/oxford-english-dictionary-change-oxford-

english-dictionary-s-archaic-definition-of-an-

accountant?utm_source=twitter&utm_medium=social&utm_campaign=change_oed

CNBC (2019). 40% of A.I. start-ups in Europe have almost nothing to do with A.I., research finds.

Retrieved: https://www.cnbc.com/2019/03/06/40-percent-of-ai-start-ups-in-europe-not-related-to-

ai-mmc-report.html

Financial Institutions

Group

References

EY (2018). How artificial intelligence will transform the audit. Retrieved:

https://www.ey.com/en_gl/assurance/how-artificial-intelligence-will-transform-the-audit

Financial Times (2018). PwC’s failure to spot Colonial fraud spells trouble for auditors. Retrieved:

https://www.ft.com/content/c2cc45d6-f1f6-11e7-b220-857e26d1aca4

Forbes (2016). How Artificial Intelligence Will Transform The Delivery Of Legal Services.

Retrieved: https://www.forbes.com/sites/markcohen1/2016/09/06/artificial-intelligence-and-legal-

delivery/#20c428cf22cd

Forbes (2018). How Much Data Do We Create Every Day? The Mind –Blowing Stats Everyone

Should Read. Retrieved: https://www.forbes.com/sites/insights-kpmg/2018/10/19/artificial-

intelligence-real-breakthroughs-the-practice-and-promise-of-ai-in-auditing/#1aee86126c10

https://www.forbes.com/sites/bernardmarr/2018/05/21/how-much-data-do-we-create-every-day-

the-mind-blowing-stats-everyone-should-read/#7ef6f8cf60ba

Forbes (2018b). Artificial Intelligence, Real Breakthroughs: The Practice And Promise Of AI In

Auditing. Retrieved:

Huffpost (2015). Disrupting Today’s Healthcare System. Retrieved:

https://www.huffpost.com/entry/disrupting-todays-healthc_b_8512200

Financial Institutions

Group

References

•Journal of Accountancy (2011). Special Focus Report: Going Paperless. Retrieved:

https://www.journalofaccountancy.com/news/2011/jul/july2011goingpaperless.html

•Markets and Markets (2019). Artificial Intelligence Market by Offering (Hardware, Software, Services), Technology (Machine

Learning, Natural Language Processing, Context-Aware Computing, Computer Vison), End-User Industry, and Geography –

Global Forecast to 2025. Retrieved: https://www.marketsandmarkets.com/Market-Reports/artificial-intelligence-market-

74851580.html

•Riedl,Mark (2017). Human-Centered Artificial Intelligence. Retrieved: https://medium.com/@mark_riedl/human-centered-artificial-

intelligence-70b019f956d1

•Spangler, William et al. (pending publication). Educating Business Students for the Age of Intelligent Machines: A Framework

for On-Line AI-enabled Learning.

•TechTerms (no date). Data. Retrieved 4/30/2019: https://techterms.com/definition/data

•The CPA Journal (2017). Big Data in Business Analytics: Implications for the Audit Profession. Retrieved:

https://www.cpajournal.com/2017/06/26/big-data-business-analytics-implications-audit-profession/

•The Globe and Mail (2017). Supreme Court says Livent auditors liable but sets conditions. Retrieved:

https://www.theglobeandmail.com/report-on-business/supreme-court-says-livent-auditors-liable-but-sets-

conditions/article37393018/

•The Guardian (2017). BT loses almost £8bn in value as Italy accounting scandal deepens. Retrieved:

https://www.theguardian.com/business/2017/jan/24/bt-loses-7bn-in-value-as-italian-accounting-scandal-deepens

•The Guardian (2018). KPMG to fine staff £100 for late time sheet. Retrieved:

https://www.theguardian.com/business/2018/dec/20/kpmg-to-fine-staff-for-late-time-sheets

•Wikipedia (no date). Optical character recognition: Retrieved: https://en.wikipedia.org/wiki/Optical_character_recognition

•Wikipdedia (no date). Cloud computing. Retrieved: https://en.wikipedia.org/wiki/Cloud_computing

•Wikipedia (no date). Blockchain. Retrieved: https://en.wikipedia.org/wiki/Blockchain

•Wikipedia (no date). Big data. Retrieved: https://en.wikipedia.org/wiki/Big_data

•Wikipedia (no date). Machine Learning. Retrieved: https://en.wikipedia.org/wiki/Machine_learning

•World Economic Forum (2018). The Future of Jobs Report 2018. Retrieved: https://www.weforum.org/reports/the-future-of-jobs-

report-2018

56 Insight. Oversight. Foresight. SM

Thank You!

Michigan Texas Florida  North Carolina

Robin D. Hoag, CPA, CGMA, CMC

Shareholder, Financial Institutions Group

Office: (248) 244-3242

Cell: (248) 709-1270

hoag@doeren.com

2 views·56 pages

Data Analytics: Implementation PDF Free Download

Data Analytics: Implementation PDF free Download. Think more deeply and widely.

Uploaded by Michael Williams on 4/14/2026

/56

100%