Transcript Dr Richard van der Wath

Technology Systems – the Business Enabler
Richard C. van der Wath, PhD (Cambridge)
AACC 2016
17 May 2016
MYBUCKS – OVERVIEW AND KEY HIGHLIGHTS
MyBucks is a Luxembourg-based company with operations in 12 countries and is
currently listing on the DAX.
FINANCIAL PRODUCTS PORTFOLIO
PRODUCT OFFERING UNDER 3 BRANDS:
•
GetBucks: (Consumer Loans) Short-term single
payment loans and instalment loans as well as
additional value added services such as a credit
report and a budgeting tool.
•
GetSure: (Insurance Products) Credit, legal,
funeral and medical cover.
•
GetBanked: (Mobile Banking) Online savings
accounts, mobile transactions and transaction
cards.
• With more than 380 employees and over 400 sales agents MyBucks
has already disbursed more than 598,000 loans since inception
and over 2,5mil customer interactions.
• Our proprietary FinTech platforms enables fast and
automated disbursements of loans within fifteen minutes of
loan application.
o Jessie is our credit decisioning and scoring technology
platform.
o FinCloud is our fully cloud-based proprietary loan
management system.
CREDIT REPORT
•
•
•
•
Feedback on credit scores, financial history, spending habits to educate customers
Fraud detecting
Show availability of financial products
Rehabilitate blacklisted clients
FINANCIAL
BUDGETING TOOL
•
•
•
•
Understanding spending habits and
monthly liquidity requirements to
make smarter budgeting decisions
Expenses are categorized
Setting of monthly saving goals
Send warnings for projected
overspending
PAGE 1
SEPARATING FROM TRADITIONAL BANKS THROUGH FINTECH
OVERVIEW OF COMPETITIVE LANDSCAPE
Range of
financial
products
Traditional
Banks
Microfinance
Institutions
Online
Lenders
Mobile Money
& Payments
Payday
Lenders
Brick & Mortar
Online
Mobile/Digital
Distribution Channels
Source: Company analysis.
PAGE 2
BUILDING A LEADING DIGITAL BANK
KEY FACTS
•
•
OUR KEY MILESTONES
Our product offering encompasses shortterm single payment loans and installment
loans, as well as insurance products and
online savings accounts, mobile transactions and transaction cards.
FINTECH
PIONEER
PRODUCT
BROADENING
•
•
Geographical
expansion
including Europe
distribute our products at a highly
competitive cost;
rapidly scale up our business,
effectively manage our credit risk.
How?
Answer: Technology systems!
TRANSFORMATION
Opportunity Banks:
Deposit taking
institutions
Our proprietary credit decisioning and
scoring technology and self-learning
algorithms enable us to:
•
COUNTRY
EXPANSION
Development of
proprietary
Fintech platform
Start of mobile
and online
lending
Start of
insurance
business
*
*
*
Note: Acquisition refers to the acquisition of six financial institutions from Opportunity International Inc.; * No operations, only holding and financing subsidiaries
PAGE 3
OUR PROPRIETARY FINTECH PLATFORM
•
•
•
•
FinCloud, our fully cloud-based proprietary loan
management system, enables us to manage credit
risk, our loan book portfolios, and efficiently serve
our customers securely, via the internet, mobile
phone and telephone, in real time.
Jessie, our proprietary credit decisioning and scoring
technology platform, is based on artificial
intelligence (AI) algorithms . It considers a number of
factors, such as the customer’s behavioural,
transactional and credit bureau data, as well as
employment information.
Watson uses AI algorithms (e.g. clustering and
classification) to calculate a client's fraud score
based on how close that client's online behaviour
matches past fraudulent behaviour of known
fraudsters. Clients with high fraud scores are flagged
for further review.
Redundancy provided by tier 1 Data centers in Africa
and Europe.
•
Live reporting systems provide
management with real-time
company performance.
•
Security is provided through various
encryption technologies.
•
Our customer and internal processes
are seamless as we fully integrated
third party platforms into our
systems (e.g. several govern-ment
platforms across the continent)
•
BR.NET is MyBucks banking platform
for microfinance banks in the group.
•
TERMENOS is the banking platform
for the retail banks which includes
the Opportunity operations.
PAGE 4
THE MYBUCKS UX ECOSYSTEM
PAGE 5
MOBILE INTERACTION
Log-in
Dashboard
Sign-up
Product
Selection
Credit Report
Budget Tool
PAGE 6
BACK-END OVERVIEW
PAGE 7
A.I. BASED CREDIT SCORING AND FRAUD DETECTION
ARTIFICIAL INTELLIGENCE (AI) BASED CREDIT SCORING
WHAT IS AI & MACHINE LEARNING?
• AI is an advanced form of data analytics that can find complex
patterns in data using mathematical and statistical methods.
• Since AI algorithms are self-learning from data, this type of AI is
also sometimes referred to as ‘Machine Learning’.
HOW CAN AI BE USED FOR CREDIT
SCORING?
• MyBucks developed an in-house AI
algorithm that can learn client
behavioural patterns from historical
data to build a model that can predict
the Probability of Default (PD) per
client, per product.
• Since this prediction is a probability (a
value between 0 and 1) we can align
this to our chosen risk appetite.
WATSON – A.I. FRAUD SERVICE
KEEP YOUR FRIENDS CLOSE AND YOUR ENEMIES EVEN CLOSER…
• The fraud service at MyBucks creates a barrier to entities prying on our loan application
process – entities wishing to create many different seemingly unique registrations,
• The fraud service deals with two types of fraudsters:
o Opportunists – many and unprofessional, ‘easy’ to stop, hit-and-run criminals,
but with low impact
o Syndicates – fewer but professional, difficult to identify, targeted attempts
with potentially high impact
• Fraudulent client registrations are created, ready to apply for loans at a later stage in
an orchestrated attempt to defraud the company,
• Fraudsters attempt as far as possible to mimic honest clients,
• The fraud service attach a quantitative value to each registration, on the basis of
similarity with every other client,
• The fraud service analyses registration detail and logon pattern in real time to attach a
suspicion number to a registration,
• Any application breaching this level of suspicion is passed on to a bridged process,
where the application is held for further review,
• Where possible biometric information such as online face recognition will add more
value to the fraud service,
PAGE 8
JESSIE – A.I. CREDIT
SCORING
• Jessie only influence product offering (i.e. the maximum amount
offered).
• Allows setting of loan term specific risk appetite.
PAGE 9
WATSON – A.I. BASED FRAUD DETECTION
AI BASED FRAUD DETECTION
FRAUD
•
•
•
Online – who are you dealing with?
Raise a flag (don’t reject outright . . .)
Call for human intervention.
ANALYSIS OF SIMILARITY (ANOSIM)
•
Does a ‘similar client’ already exist?
•
What is similar (equal is easy)?
•
ID numbers, email, IP address, facial recognition
•
How much work / effort / energy is needed to change something
into something nearly the same?
PAGE 10
A.I. BASED CREDIT SCORING & FRAUD DETECTION OUTCOME
AI BASED CREDIT SCORING & FRAUD DETECTION - OUTCOME
IMPROVED RISK MANAGEMENT
•
Credit offering automatically set based on client’s predicted risk profile and an adjustable risk appetite.
•
Pre-emptive managing of fraud risk.
.
TAILOR MADE CLIENT OFFERING
Overall client experience improves if loan amounts and installments are intelligently matched to what clients can
comfortably afford.
.
PAGE 11
The End
[email protected]