how to leverage big data to help finding fraud patterns & revenue assurance
DESCRIPTION
A joint presentation by Mobitel Sri Lanka and cVidya. Delivered on the Telecoms Fraud Management and Revenue Assurance World Summit 2014 in SingaporeTRANSCRIPT
© 2014 – PROPRIETARY AND CONFIDENTIAL INFORMATION OF CVIDYA
How to Leverage Big Data to Help Finding Fraud Patterns & Revenue Assurance
Sandagomi Jeewapadma, GM Enterprise Risk Management, Mobitel Sri Lanka
Amit Daniel, EVP Marketing & BD, cVidya
2
About Mobitel - Sri Lanka
Sri Lanka is a small island nation with a population of 21 million According to regulator statistics, by September 2013:
2013 Subscriber
(Mn) Subscriber
share
Mobitel 5.0 25%
Dialog 7.5 38%
Etisalat 4.5 23%
Other 2.8 14%
Mobile Subscriber Market Share
Number of Mobile Subscribers 20,234,698
Mobile Subscription per 100 people 98.78
There are five Mobile Operators, two of those have launched 4G LTE services to the market (Mobitel & Dialog) Mobitel, a wholly owned subsidiary of Sri Lanka Telecom, is a mobile and broadband service provider In Sri Lanka
Mobitel 25%
Dialog 38%
Etisalat 23%
Hutch - 5% Airtel - 9%
3
A leading supplier of Revenue Analytics solutions to communications and digital service providers
Founded: 2001
300 employees in 15 locations worldwide
Deployed at 7 out of the 10 largest operators in the world
150 customers in 64 countries
Processing 2.45 Billion subscribers in deployments globally
Saving over $12 Billion to providers annual revenue
Partnering with world leading vendors
About cVidya
4
BUSINESS ANALYTICAL LAYER
BUSINESS GROWTH
BUSINESS PROTECTION
Transformation Assurance
Fraud Management
Revenue Assurance
Marketing Analytics
Sales Performance Management
BIG DATA PLATFORM
Data collection
Aggregation
Enrichment
DWH
CRM Mediation
ERP
IP&DPI Probes
Switch Billing
Order & Provisioning
DATA SOURCES
Domain Expertise
Education Center
Professional Services
Business Consulting
Turning your DATA into VALUE
5
Market Trends
Source – GSMA Global cellular market trends and insight – Q4 2013
6 6
An Entire New Ball Game
7
Fraud & RA Units Must Process & Control Huge Amounts of Data
From info sources that did not exist before Extensive use of external sources e.g., social networks Need for cross analysis of non-associated sources of info Including a new set of risks and threats to be identified & controlled Entails a whole new terminology to master and areas to cover
8
Mobitel Case Study
9
New RA & Fraud Challenges
Industry definitions are rapidly changing
New complex data services are being populated across operator service offerings
Shift from data pipe provider role to content integrator position
Complex partnerships on SLA-driven and revenue-sharing basis
New technologies and business models (LTE, Mobile Money, rich communication services etc.)
Proactive identification of revenue leakages, risk mitigation, and fraud management is essential
New set of skills and capacity required for RA & FM staff
10
Process in Selecting RA & Fraud Solution
Mobitel RA & FM function was newly set up and required visibility & control of the entire revenue map in terms of revenue leakages and fraudulent activities
Creating internal capacity was also a mandatory requirement, to be executed in parallel
Mobitel invited leading players in the domain (based on Gartner’s Magic Quadrant)
Comprehensive technical evaluation process, qualified by cVidya for RA & FM solution
Inclusion of organizational key risks & revenue sources and defining correct control points & KPIs are essential at the beginning of the project
11
Mobile Money – Mobitel
New partnership with banks, merchants New regulatory authorities (central banks) New risks & controls on KYC and money laundering threats
12
It’s Complicated… So Much to Check!
GGSN IP SGSN
HLR
BSC/RNC
MSC
SMSC
Gateway Router
Service Platform & Portal
AAA RADIUS
Mobile Network
Customers Agents Merchants
Bank ATMs Agents Merchants
Customers Agents Merchants
CRM
Billing: • Postpaid • Prepaid
Reports • Banks • Agents • Merchants • Others
PSDN
www
Secured Network
13
LTE Challenges
Consumption or service based is much more complex than transport based charging
New service requirements with Shorter Time to market
Complex Price plans
Quality Of Service based rating create new challenges for verification and re-performance
Multiple charging policies in the same session
So…
Do we measure usage correctly?
Are we applying the appropriate policy?
Are we charging according to the appropriate policy?
13
14
It’s Complicated… So Much to Check!
EPC
e-NodeB e-NodeB
RAN
S-GW
MME
P-GW
HSS
PCRF SPR
ePDG
PDF CSCF
AS
MGW
IMS
MGCF
OFCG OCG
Wholesale Billing
CRM
Postpaid Billing
e-NodeB
www
PSTN/PLMN
PCEF
Service Configuration
Portal
Configuration
Usage
RBA
15
Big Data Analytics for Fraud Management
Using Deep Packet Inspection (DPI) and Pattern Matching is highly effective for:
Identifying malicious calls & applications in real time
Detecting abnormal service consumption
Detecting subscriber frauds
Mobile Money related Frauds (Phishing attacks)
Detecting Tethering of Smart Phones
Detecting Proxy Services
Achieving visibility on OTT services
16
Using DPI to identify Fraud
Tethering performed in a commercial manner is considered to be an abusive operation and impacts the telecom operator in several ways:
– Affecting the network planning and causing overloads
– Could force the operator to invest in expanding his network – Harming the user experience of other legitimate users
www
ISP Backbone
Legitimate connections Non-Legitimate connections
Abusive tethering operation
17
Quicker, Richer, Better Cross analysis of non-correlated sources
Accurate, fast & intelligent insights
Reduction of time & investigation resources
Larger retention - storing for longer periods of time
Enabling RA & Fraud units to provide services and leverage capabilities for other non-fraud activities
When Big Data Meets RA & Fraud
THANK YOU! www.cvidya.com