EY: Why graph technology makes sense for fraud detection and customer 360 projects
•Transferir como PPTX, PDF•
0 gostou•195 visualizações
This document discusses why graph technology is ideal for customer 360 and fraud detection projects in the insurance industry. It provides an overview of graph use cases in banking, insurance, and capital markets including for customer 360, fraud detection, and knowledge graphs. It then discusses challenges insurers face with siloed data and lack of a unified customer view. Implementing a customer graph allows linking diverse data sources to create a complete view of customers and their relationships to enable context-based decision making and analytics.
Recomendados
Recomendados
Mais conteúdo relacionado
Mais procurados
Mais procurados (20)
Semelhante a EY: Why graph technology makes sense for fraud detection and customer 360 projects
Semelhante a EY: Why graph technology makes sense for fraud detection and customer 360 projects (20)
Mais de Neo4j
Mais de Neo4j (20)
Último
Último (20)
EY: Why graph technology makes sense for fraud detection and customer 360 projects
- 1. Page 1 Why graph technology is ideal for Customer360 and fraud detection
- 2. Page 2 1. Introduction & Background 2. Graph Use Cases in Banking, Insurance and Capital Markets 3. Customer360 and Fraud in Insurance 4. How to get up and running AGENDA Page 2
- 3. Page 3 Introduction Kenneth Nielsen, PhD Director Kenneth.Nielsen@dk.ey.com +45 25 29 32 61 Kasper Müller Manager kasper.mueller@dk.ey.com +45 25294765 Ivana Hybenova Senior Consultant ivana.hybenova@dk.ey.com +45 25295065 Sebastian S. Bennetzen Consultant sebastian.bennetzen@dk.ey.com +45 25296373 • Kenneth is a Director in FS Technology Consulting where he heads up the Data and Analytics area. Prior to joining EY in 2017, he worked 3.5 years as a senior quant in Nordea Markets with a high focus on optimizing the use of data across teams. • His team focuses on big data analytics, network analysis and streamlining/automating processes using a combination of process mining, machine learning and programming. Kenneth has been delivering projects revolving around these topics in both global banks and most recently in Danske Bank. • Kasper is a Manager in FS Technology Consulting and in practice, he is working as a data scientist and programmer, designing and building mathematical and programmatical models and solutions. • His experience counts DevOps, Machine Learning, Cloud Architecture, and implementing scalable solutions using Docker and Kubernetes. • Kasper infuses graphs on the back of EY-developed solutions such as the GDPR Scanner and holds both the Certified Developer and Graph Data Science certification • Ivana has background in statistics and has been working as a data analyst and scientist for over 4 years. • She currently leads the organization of the EY Graphathons as well as two internal graphs projects on customer analytics and AML • Ivana is Neo4j Certified Professional and has also Neo4j Graph Data Science certification. • Sebastian is a consultant in FS Technology Consulting. He has obtained a masters degree in Mathematics and Economics with a focus on graph theoretic studies from SDU. • At EY, Sebastian has been working on fostering Graph Analytics by hosting multiple Graphathons and contributing to internal graph-based projects. • Sebastian has completed the Neo4j Certified Professional and Neo4j Graph Data Science certification. Why graph technology is ideal for Customer 360 and fraud detection
- 4. Page 4 Background … Why graph technology is ideal for Customer 360 and fraud detection 5 4 3 2 1 5.5 years ago • First encounter 3.5 years ago • Decided to have a second look 1 year ago • Graph Community (Nordic) setup 2.5 years ago • Established the first small in-house EY group • Initial relationship with Neo4j TODAY • EY Graphatons • Part of Global Graph community
- 5. Page 5 …. and a warning Why graph technology is ideal for Customer 360 and fraud detection
- 6. Graph Use Cases in Banking, Insurance and Capital Markets
- 7. Page 7 Banking, Insurance and Capital Markets Why graph technology is ideal for Customer 360 and fraud detection Enterprise Knowledge Graph Customer 360 Financial Crime Identity Access Management
- 8. Page 8 A B C D E A B C D E One-to-Many Relationships Across Many Entities Wide Data Complex Data Hierarchical & Recursive Data Many-to-Many Relationships Nested Tree Structures Recursion (Self- Joins) Deep Hierarchies Link Inference (If C relates to A and A relates to E, then C must relate to E) Node Similarity Hidden Data Legacy Data Frozen Data Legacy SQL Systems Data Lake Fact Tables Graph Data Science - Machine Reasoning A C E If your data looks like this, you should be thinking graph Why graph technology is ideal for Customer 360 and fraud detection
- 9. Page 9 A B C D E A B C D E One-to- Many Relationship s Across Many Entities Wide Data Complex Data Hierarchical & Recursive Data Many-to-Many Relationships Nested Tree Structures Recursion (Self- Joins) Deep Hierarchies Link Inference (If C relates to A and A relates to E, then C must relate to E) Node Similarity Hidden Data Legacy Data Frozen Data Legacy SQL Systems Data Lake Fact Tables Graph Data Science - Machine Reasoning A C E Enough with the abstract Customer Product Basic form Tegningsgrundlag Conditions Status Risk savings Payment 1 Payment 2 D10006 Traditional 211 D5 1 3 S PU P D10007 Traditional 810 D5 1 4 R PU P D10008 Traditional 810 D3 1 4 R PU P D10008 Traditional 810 D5 1 4 R PU P D10009 Traditional 810 D5 1 4 R PU P D10009 Traditional 840 D5 1 4 R PU P D10010 Traditional 810 D2 1 4 R PU P D10010 Traditional 840 D2 1 4 R PU P D10010 Traditional 810 D3 1 4 R PU P D10010 Traditional 840 D3 1 4 R PU P D10010 Traditional 810 D5 1 4 R PU P D10010 Traditional 840 D5 1 4 R PU P D10011 Traditional 115 D1 1 1 R PP E D10011 Traditional 185 D1 1 1 S PP E Why graph technology is ideal for Customer 360 and fraud detection
- 10. Customer360 and Fraud Case
- 11. Page 11 Inability to recommend Next Best Action (NBA) Non-optimized fraud identification and actioning capabilities Lack of full view of customers and agents Silo-ed legacy systems Obsolescence of Enterprise Data Warehouse Fast changing customer needs Primarily broker-mediated market Recent fraud trends - Deepfakes Increased manual processing Reporting vs recommending Reactive rule-based policies Operations at scale Caused by Insurers are facing some key challenges which impacts growth ! Why graph technology is ideal for Customer 360 and fraud detection
- 12. Page 12 UNIFIED VIEW OF THE CUSTOMER Marketin g Sales Policy Claims Contact Centre Broker External Data Demogra phics A unified Customer 360° view enables: • Data-driven, customer-centric experiences • Efficient and automated sales & marketing • Improved compliance and better underwriting through fraud detection • Consistent view of operational metrics across business segments • Improved decision-making based on more reliable reporting Customer 360° View in an Insurance Company Why graph technology is ideal for Customer 360 and fraud detection
- 13. Page 13 Example Schema: Insurance Agent 360° Marketing UNIFIED VIEW OF THE CUSTOMER Demo graphic Policy Claims Contact Center External Data Broker Sales Why graph technology is ideal for Customer 360 and fraud detection
- 14. Page 14 Why graph technology makes sense for fraud detection and customer 360 projects in insurance Integrating data into the graph and building analytics and BI-layers on top to enable rapid extraction of cross-LOB features on a customer for context-based decision making Integrating data into the graph and building analytics and BI-layers on top to enable rapid extraction of cross-LOB features on a customer for context-based decision making Rapidly test and operationalize new analytical capabilities Who are our customers? What drives a customer to make a buy decision? How to understand different customer behaviours? How to get right and up-to- date information about every customer? How to create effective risk policies? We want to understand… Identify & ingest multiple data sources All data is aggregated and linked together in the graph that provides an entity centric view of all customers, products, and merchants Link and maintain graph database Create new data assets and signals Key components of the Customer Golden Profile Customer Agent Product Master Data Quote Policy Claims Customer Journey Data Risk & Compliance Insured Asset Third-party External Data Build a complete view of customers’ relationship with businesses Identify key data elements and customer behaviors within and across Lines of Business Develop customer 360 level attributes that are predictive of customer behaviour Example Capabilities Predict Churn Personalise product bundling Optimise discount via agent effectiveness Predict conversion in sales cycles Predict effectiveness of cross-sell & up-sell schemes Predicting fraud triangles Effective Chatbot for Contact Centre activities Customer Golden Profile will create cross-LOB data assets to help answer key strategic questions
- 15. Page 15 Increase Cross- Sell and Upsell Increase Retention Increase Customer Satisfaction Reduce Cost to Acquire and Service Reduce Fraud And how to measure them? Value(DKK) and Volume(#) of policies sold to existing customers in a year Measure what matters . . . Annual customer churn rate across and within LOB Average of CSAT score and Annual NPS score Average time-to- resolve at Contact Centre Direct and Indirect expenses by Customer Journey milestones Loss ratio and combined ratio Straight- Through- Processing policies How Graphs add value to the insurance business Why graph technology is ideal for Customer 360 and fraud detection
- 16. Page 16 Sales / Marketing Customers are not always “price sensitive” but “value sensitive” Referral programs are effective along with product bundling Agent is the “influencer” but customers always validate the information online Discount optimisation based on “influence capability” of the agents Risk & Compliance Increased risk exposure due to “serial” entrepreneurs (a.k.a habitual offenders) Common elements between claims - like garage, doctor, 3rd party in car & liability insurance, etc. Loss of opportunities from traditional rule-based risk policies – E.g., a young driver is not always the riskiest driver Some of the interesting insights were Why graph technology is ideal for Customer 360 and fraud detection
- 17. Page 17 Can’t we just do this in an RDBMS? Customer Golden Record • Slow execution • Faster execution EDW Schema Graph schema Customer Golden Profile Before After Source systems LOB - System 1 LOB - System 2 LOB - System 3 Agent Quote Name Address Phone Policy Claims Broker Why graph technology is ideal for Customer 360 and fraud detection
- 18. Page 18 Many companies today utilize Customer Graphs: To support the demands of the digital business, enterprise architects must consider how best to link large volumes of complex, siloed data... Graph databases are a powerful optimized technology that link billions of pieces of connected data to help create new sources of value for customers and increase operational agility for customer service. – Forrester Zurich Large online shopping site These challenges have been successfully solved using graph databases “ Why graph technology is ideal for Customer 360 and fraud detection ”
- 19. Sub-header slide Why graph technology makes sense for fraud detection and customer 360 projects in insurance How to get up and running
- 20. Page 20 Neo4j development generally begins with a graph model designed to answer a set of questions, and then data is imported to populate the model. The model is tested and adjusted as needed. Iterative, Agile Solution Development Approach Adjust if data is missing ? Start with a question Build or extend model using domain expertise Source data to populate model Import data Build query Analytics Adjust if needed to answer question Why graph technology is ideal for Customer 360 and fraud detection
- 21. EY | Building a better working world EY exists to build a better working world, helping to create long-term value for clients, people and society and build trust in the capital markets. Enabled by data and technology, diverse EY teams in over 150 countries provide trust through assurance and help clients grow, transform and operate. Working across assurance, consulting, law, strategy, tax and transactions, EY teams ask better questions to find new answers for the complex issues facing our world today. EY refers to the global organization, and may refer to one or more, of the member firms of Ernst & Young Global Limited, each of which is a separate legal entity. Ernst & Young Global Limited, a UK company limited by guarantee, does not provide services to clients. Information about how EY collects and uses personal data and a description of the rights individuals have under data protection legislation are available via ey.com/privacy. EY member firms do not practice law where prohibited by local laws. For more information about our organization, please visit ey.com. 2023 EYGM Limited. All Rights Reserved. ED None This material has been prepared for general informational purposes only and is not intended to be relied upon as accounting, tax, legal or other professional advice. Please refer to your advisors for specific advice. ey.com
Notas do Editor
- 2) Use cases, data structures that spell graph 3) Based on work done at an insurance company in Belgium with global reach – what are the data points you want in there? What could a schema look like in insurance? Couldn’t we just do this in RDBMS? Outcome of project and how to measure Not a very technical talk, but happy to discuss technicalities afterwards.
- Who am I? Team – 25 Three with me today – Kasper, Ivana and Sebastian.
- Spent past three years going down the rabbit hole of graphs Cross-Nordic graph community We host graphathons Ask Kasper, Ivana and Sebastian
- Who am I? Team – 25 Spent past three years going down the rabbit hole of graphs Cross-Nordic graph community Three with me today – Kasper, Ivana and Sebastian We host graphathons
- Takeaway – you don’t start big, you start with a specific problem! X360°: Customer 360° / KYC Agent / Financial Professional 360° Product Recommendations / Next Best Action EDF: Data Lineage / Dependency Data Exposure Process: Customer Journey Origination / Underwriting Portfolio Management Risk: AML/Fraud Detection Risk / Compliance Identity / Access Management System Dependency / Points of Failure
- It is often a mix between the data structures Requires many entities (e.g., many SQL tables, 360° views) Involves recursion (e.g., SQL self joins) Has complex, potentially colliding, hierarchies (e.g., SQL 1 to many, many-to-many) Based on informatics of the relationships themselves (e.g., collaborative filtering shared relationship counts, shortest path segment summations for wayfinding, cost/time minimization for supply chain, money flows for finance) Requires mapping, direct or indirect across data sources (e.g., data lake unification) Demands fast query results (e.g., digital applications, search)
- Another example is when you have holes (i.e. sparse) in your dataset
- Inorganic growth => 5 ERP systems. Graph for resolution of entities and products Deep fakes – document falsification and identity theft Reactive: bundling not marketed pro-actively. After the graph project became Amazon of insurance
- Talk around the circle from Broker to External Data Questions related to policy provide further demographic data Why did they want 360 view? What kinds of data do you put in there? It is not rocket science! Internal database Customer demographics (master) Quotations generated Policy Claims Payments data Broker Call centre interactions (enquiry, complaints, requests) Web portal interactions (online transactions – policy renewals, FNOL, claims, etc.) Social media interactions (enquiry, complaints, requests, feedback) External database Companies database (non-individual customers) Address database – entity resolution Point of interest – OSM GIS data Weather data Property registration, Marine vessels, Car registration data A unified view of the customer is foundational to a successful digital transformation. The customer 360° view is derived from customer, product, sales, marketing, support and web data. Data is ingested & cleaned in a data lake, unified & analyzed in a Knowledge Graph, and mobilized via API microservices.
- Provides the agent a 360° view of customer activity Connects historical data (policies, claims, quotes) with real-time interactions (customer support, web events, mobile) Handles householding complexity Supports customer analytics and next-best-action product recommendations
- X-LOB can’t be stressed enough! Leading your graph-projects with tangible questions makes things easier, but there will often also be a lot of value (and questions generated) from just starting to view your data in a graph! Some customers may be both private customers and have their company be a client – not all insurance companies pick up on this , at least not realtime. What if the ownership structure around a company changes – someone enters the ownership structure who also owns other companies that are not currently customers. Is that a threat or an opportunity?
- Example: automation of investigations in fraud and money laundering => save time, save money, catch more bad guys. The best foundation is a graph database.
- Have an optimized webpage (customer journey on webpage is also a graph) CVR data => are the owners of your company customers involved in other companies? Could they be influencers? Garage or garages. Incorporating e.g. CVR data you can start to see if there are patterns in who submits claims and/or where claims are “fixed”. You can – in principle – start to do adverse media screening. SO, why is graph tech ideal for Customer360 and Fraud?
- When is Graph superior to Relational DB – it depends… Flexibility in adding third party data, e.g. CVR, weather data, BBR data, etc. Cypher>SQL – story from DB At the end of the day it is about picking the right tool for the task! Graph can add value in any environment where: - Data is interconnected and relationships matter - Data needs to be read and queried with optimal performance - Data is evolving and data model is not always fixed and pre-defined
- The question: how many customers do I have that are n steps away from e.g. a known fraudster? How do different autoshops rank in terms of usage for specific cases? Do we have private customers who are company owners? What about the people they co-own with? Gather the data necessary. Start right, not small – don’t go overboards! Do PoC and communicate expected value to the business. Work in cross-functional teams.