Business Rules Management Systems Managing Complex Decisions in

Business Rules Management Systems Managing Complex Decisions in your Organisation Simply

Richard J. D. Collard Senior Industry Lead WW Business Development SME - Transactional Fraud, AML & Risk richard. collard@uk. ibm. com +44 7917 612896

Business Rules Management System - What Is It? ? !!!!!! BRMS - more than a Rules Engine Enabling & Complementary Component & Technology In SOA, BPM or Stand-alone Empowers Business-users Provides Ability to make Complex Decisions – Consistently, Simply and Quickly • Externalises & Makes Business Logic Available to Business-users • Hard-coded Business Logic becomes redundant • Benefits for IT and Lines of Business • • •

Why use Business Rules with BPM? Validation Rules Eligibility Rules Business Rules • Automate key decisions in Business Processes by Externalising Business Logic • Simplify the Business Process structure • Are invoked from the Business Process as standards-based Services • Are quicker to modify than the Business Process orchestration Pricing Rules

Engage and Support Separate Life Cycles § BPM and BRMS do not typically share the same velocity of change. § Business rules typically require more frequent and more rapid change cycles than a process.

Combining BPM & BRMS BPM Alone BPM + BRMS Decision Support Business Decision Making Human Automated / Assisted Decision Maintenance Process re-engineering Rules updates Decision change cycle Months Days Decision Ownership IT IT + Business Decision Sharing Cut and Paste Central Repository / TDS Scalability < 50 Rules (when rules enable) Thousands Supported decisions Routing 6 • Business Decisions • Task allocation • Dynamic routing 6

Traditional Approach for Managing Decision Change Business rules are crucial to operational systems, and they change over time. The traditional (ad hoc) approach of dealing with rule changes leads to… Where Business Rules Typically Exist Applications People Documents Processes M Reduced organizational agility M Reduced employee productivity M Increased load on IT Issues • Rules are hidden in code or isolated within the organization • Changes are hard to track and maintain over time • Rules used by systems have to be programmed and require IT resources • Duplication and multiple versions of the same rules • Lack of auditability, traceability • Decision changes cannot be easily tested or simulated 7

Manage and Automate Decision Logic with BRMS § Eliminate decision silos § Make decision logic accessible to Business and IT § Implement fine-grained, context-specific logic Business Rule Management System Where Business Rules Typically Exist Rules are Defined, Analyzed and Maintained User Tools Applications Rules are Stored and Shared Documents Rule Repository People Rules are Deployed, Executed and Monitored Processes Rule Server 8

Business Rules Made Simple § Customizable vocabulary specific to your organization, industry, application etc…. § Supports language localization § Integrates with external data sources e. g. list of countries § Drop-down lists for customized domain data § Templates facilitate new rule creation

Flexible Rule Management for Business Users Graphical representations Text-based (point-andclick + guided editing) Rule Solutions for Office

IBM ILOG JRules BRMS Product Components Business Environment Technical Environment Rule Developer Rule Studio Deploy Eclipse-based Rule IDE rules Rule Service design & development Decision Validation Service Testing & Simulation framework Rule Repository Rule Manager Deploy rules Rule Team Server Web-based Application for Rule Maintenance Production Environment System Administrator Rule Execution Server Single node or cluster (J 2 SE & J 2 EE compliant) Web-based console to monitor & control rule execution

Benefits of a BRMS Approach What does it enable? • Reduce time and resources required to deploy changes • Author and maintain rules using non-technical language • Express decision logic with increased precision • Make decisions based on specific context • Increase decision automation • Improve visibility and understanding of how decisions control systems What is the value? BRMS § Lower maintenance costs; respond quickly to change § Business experts can manage and validate decision logic § Increase profitability of product, pricing and promotional offerings § Customize decisions when possible, standardize if needed § Improve process efficiency § Ensure compliance; enable sharing/re-use of decision logic

Precise, Automated Process Decisions with Web. Sphere ILOG BRMS Horizontal: Best/appropriate price, Cross-sell/ up-sell recommendations, Loyalty promotions, Exception identification, Risk/fraud assessment, Straight-through processing approvals Insurance • Claim • Validation • STP approval • Exception routing • Policy/Underwriting • Eligibility • Risk • Pricing • Annuity • Recommendation • Commissioning • Payout calc. 13 Banking • Loan • Eligibility • Risk • Pricing • Account • Cross-sell • Fraud/Alerts • Credit Card • Mktg Offers • Fraud • Credit limit Healthcare Government • Patient Care • Drug interaction risk warnings • Follow-up alerts • Member • Services recommendation • Eligibility • Benefit calculation • Provider • Patient eligibility for services • Benefits • Eligibility • Calculations • Tax Payer • Classification • Audit flagging • Citizen • Program(s) recommendation Energy/Util. /Telco • Land/Permits • Conveyance processing • Contract compliance • Service Mgmt • Service prioritization • SLA alerts • Maintenance alerts • Order configuration

ILOG BRMS customers in payments Use case Pricing Customer Description • JRules-based pricing engine for all payments products • Differentiated pricing and complex bundle pricing • Reduction of time-to-market for pricing policies from 6 months to 10 days • ING Belgium considers its business as unique and cannot manage its pricing policies using a package Selection of clearing channels – Minimizing C&S costs • Consistent selection of least-cost channel (by consolidating all routing decisions in a centralized decision service) • Rapid adaptation to price changes (e. g. new pricing scheme of channels, “on us” settlement channel after acquisition) Semantic validation of messages • Increase straight-through processing • Separate valid messages that can be processed in STP from messages that need exception handling

ILOG BRMS customers in payments Use case Customer • Increase straight-through processing (countryspecific validation rules) • Integration with EPP-based payments hub • Other use cases being analyzed: customer-specific conditions for corporates Validation of account numbers Anti-Money Laundering Card fraud detection Description Top-10 US Bank • Rule service performing country-specific AML checks • Integration with EPP-based payments hub • Standard-based (ISO 20022) service definition • Fraud patterns captured as business rules to react quickly to new fraudulent behaviors

ILOG BRMS customers in card and online payments Use case Customer Description • 150 M transactions processed in a 8 hour time window • Shorter time-to-market for new products • Specific rules for various regions Message validation, calculation of merchant fees & internal routing Clearing and Settlement System • Expose business policies for validation, interchange and pricing • Speed of change; defect analysis, handling late change • Flexibility; ability to re-order rules or change flows quickly • Quicker learning curve for new starters Merchant-level fraud detection • Merchant-level fraud detection provided as a valueadded service to banks • Business rules as a way for TSYS to exploit value by extracting knowledge from huge volume of card transactions Online fraud detection Global e-Auction • Fraud patterns captured as business rules to react quickly to new fraudulent behaviors

Case Study: European P&C Insurance Company Using BPM and BRMS Together § Chose the claims process as the best target for Nordic-level unification and automation § Combination of Web. Sphere Process Server and Web. Sphere ILOG JRules was based on the positive results from a test project in Finland: ü Decreased the cost of claims incurred ü Increased of customer satisfaction through faster turnaround ü Substantial savings of employee work § Uses of BRMS in this solution: Claims submission validation Decision automation for liability and compensation Payment calculation Straight-through processing determination Manage and automate decision variation across countries, processes and systems § Accelerate implementation of decision changes § § § 17

Société Générale – Customer Risk Management • SG : Major French retail bank, 9 M+ customers, 2500 Branches, 13, 000 branch advisors • Branch Advisors spend 1, 5 hr each day to understand their customer’s account situation and make decisions on what to do next • They apply their own knowledge to decide which customer situation to fix first and how to fix it • There is a fairly high turnover of staff, so there is little control on this knowledge • The Bank’s Management needs to understand how many and which situations are solved, handled, and which ones may become critical • Management has identified there is a possibility to reduce risk, whilst improving the advisors’ productivity and support

Project High-level Specification • Identify risky situations on customer groups • A risky situation is – An non compliant situation that requires an action from the bank – It is generated, modified, closed upon reception of an event – It applies to an individual or to a service – It stays open until resolution • Risky situations are presented to branch advisors on a daily basis to help them – analyze the level of customer risk – assess which decisions should be taken. • Assist Branch Advisors by recommending a course of action – Explanations – Integrated rule based pre built customer letters, « click and send » – Integrated interview management – Notification of actions to other applications • Build an history of actions taken by the advisor for each customer • Provide a synthetic presentation of all open situations, by level of urgency/priority • Provide a synthetic view of open situations at the Branch level

Business case & targeted ROI • “High risk customer profile” process: Night Advisor’s working day Get explanations And actions plan 4, 000 events on 9, 000 accounts Get list of letters For a customer 400, 000 High Risk profiles about 5% of customers Get High Risk Profiles When logging in Update actions plan 13, 000 advisors 2, 500 branches • Benefits: – Manage 200 risk management profiles in 7 categories – Allow business users to manage & understand the risk profiles – Gain 1 hour/day/advisor Ø 50, 000 letters per day

Société Générale – Risk Management Situation Goals Solution Benefits 21 § Long time spent each day by branch officers to assess customer and account situations requiring actions § Complexity of cases and related actions § Inconsistent operations leading to unmanaged situations § Improve risk detection and management § Increase branch financial advisor efficiency managing risk profiles § Automate action recommendations § Service Oriented Architecture sharing 400 business rules § 7 rule services ( Batch and Transaction processing programs) detecting and managing 400, 000 high risk customer cases § Determine status and priority of the risky situations § Recommend and explain client actions to branch financial advisors § Generate personalized mails as actions ( 100, 000 to 200, 000 mails per day) § Manage 200 risk management profiles in 7 categories § Allow business users to manage and understand the risk profiles § Gain 1 hr/day/advisor (13, 000 advisors in 2, 500 branches)

BNP Paribas - Project OPRA Business Requirements • Offer and Product Repository Applications • BNPP wanted to improve its understanding and management of the multiple banking offers that were being made to customers • A - Flexible solution to allow definition of banking offers with • B - Ability to understand relationship between offers and • C - Determine conditions for appropriate, ethical sale • Requirement to allow Client Advisors to look into the product catalogue and filter by customer eligibility and profile with differential pricing tariffs applied • Need to distribute offers to operational systems – mainframe and distributed

BNP Paribas - Project OPRA Implementation • OPRA is based on IBM Info. Sphere MDM Server for Product Information Management and……. . • ……. IBM Web. Sphere ILOG JRules / BRMS • Info. Sphere MDM Server provides offer authoring and relationship management……. • …. thereby ensuring offer package compatibility, base tariff enforcement and conditional eligibility • ILOG BRMS provides the capability for BNPP to model eligibility and pricing rules • The SOLUTION allows BNPP to build / maintain a catalogue of of product / offer packages while managing collaborative processes associated with implementation / deployment of Business Rules

BNP Paribas - Project OPRA Benefits • OPRA is based on IBM Info. Sphere MDM Server for Product Information Management and……. . • ……. IBM Web. Sphere ILOG JRules / BRMS • Together providing: • Improved Time-To-Market for new products and services • Ability for non-technical users to generate and manage offers • Accurate and complete view of offer / product information • Multi-channel servicing initiatives • Personalised offerings • Evolution of offers according to market dynamics • Reduced likelihood of mis-selling

Automating decisions along a payments flow Validation Product configuration • Validation & autorepair • Validation of account numbers (countryspecific rules) • Determine product and configure payment attributes based on instructions and customer profile Approval rules SLA-based exception mgt • Customer-specific and SLA-based limits • Determine what to do next upon an exception • Automate SLA conditions 03/02/09 Routing • Least-cost routing in CSM selection • Customer preferences • CSM operational constraints, e. g. type of payments supported, cutoff times … SEPA DD • Mandate management Fee calculation • Real-time charge calculation • Monthly discount calculation • Payment bundles • Agreement and customer-specific conditions Accounting & reconciliation • Example: How to reverse a cancelled payment that is part of a batch 26

Customer Success – Simplifying the Process Customer Process Pains Loan Origination • Highly manual process • Web. Sphere ILOG BRMS to handle a • Delays arose as variety of business applications were logic for a IBM BPMrouted through orchestrated process: approvals via - Routing interoffice mail - Compliance - Scoring & rating • Lengthy approval turnaround time • Cost savings – $1. 5 M over 5 years • Fragmented systems • IBM BPM to manage with inconsistent data the application meant decisions to process end-to-end grant benefits were not consistent from • ILOG JRules to one center to another determine the eligibility of applicants • Several departments to social benefits but no clear “interface” for the citizens • Better service delivered to the citizens: fewer interactions required, clear and fair application process • Consistency of eligibility decisions: same criteria and rules applied across all agencies and centers • Transparency and auditability: eligibility rules can be reviewed and validated by non-technical users Social benefits eligibility 27 BPM - Decision Service Results • 15 days to 1 approval time reduction • Improved staff focus and satisfaction

IBM ILOG Customers. . . e-Gov Capital Markets Insurance Banking T & T etc. .

The Value of Simplification

Five Ways to Fast ROI With BRMS Cost reduction 1. Reduced application development and maintenance 2. improve payments margins 3. Improved regulatory compliance Revenue generation 4. Provide value-added services 5. Faster time-to-market

Using the Right Tools for Different Types of Rules Integration with other Components & Services Rules Types Rule Examples Where are Rules Defined? Business Event Stream Pattern-Match Event pattern: person on the CIA watch list, purchases chemicals , purchases a one-way plane ticket, uses cash. Eventstream specific Business Events (CEP) Assign loan review task to the Process Flow / Task Underwriter with the lowest number of Assignment tasks-in-queue. Process specific BPM Trigger notification / process when Process Monitoring “Response Time” KPI for Final Approval (SLAs) task exceeds maximum threshold value. Process specific BPM-BAM Decoupled from applications BRMS Variable, Complex Business Decisions and Policies 31 Best Offer (pricing and promotions), Eligibility Determination, Risk Assessment and Mitigation, Straightthrough Processing Determination Right Tool

Business Event Processing/BRMS Combined Use Case - Retail Promotional Offers • Use BEP to detect opportunity – Customer with regular history of book purchases visits website after 3 months absence – Customer browses to 2 successive book categories with no shopping cart activity Raise opportunity alert • Use BRMS to determine/decide whether to offer a targeted special, and if so, for what – Apply customer information against promotions rule-set – Determine best offer (e. g. , no shipping charges for any purchases in the next 1 hour)

Business Processing & BRMS Combined Use Case - AML Anti-Money Laundering • Use BEP to detect situation at risk – Successive near net-zero cash deposits/transfers – Absence of previous history of this pattern, or special circumstances • Use BRMS to determine if a genuine violation, no problem, or a “watch” – Evaluate customer history and events data against fraud score and regulatory policies – Evaluate overriding factors – If not an actual violation, determine likelihood that activity is suspicious set watch or investigate status

BRMS for Legacy Applications Business Rule Management System Rules are Defined, Analyzed and Maintained • • • Increased agility for management of automated business decisions in COBOL applications Improved visibility of decision logic Enables progressive application modernization strategy User Tools Rules are Stored and Shared Rule Repository New in v 7. 0 • Simplified creation of Business Object Model for COBOL applications • Enhanced ability to import, verbalize and manage the business model for the COBOL project • Additional JRules rule authoring capabilities for generation of enhanced COBOL code Rules are generated as COBOL copybook for execution in IMS, CICS, batch

JRules and Lombardi Teamworks: Standards-based Integration Map to JRules Rule Services via standard Teamworks Integration Services Rule Management Teamworks Process Communication Interface / ESB User Tools DEPLOY Rule Repository Rule Execution Server Windows/Unix/Linux Server Business Initiative Presentation to Agricorp Interface Flexibility: Ø Web Services Ø Messaging Ø Java APIs

File. Net BPM/ECM & Web. Sphere ILOG BRMS Business Rule Management File. Net Business Process Integration Flexibility • Web Services • Messaging • Java/. NET APIs Web Services / APIs Messaging Queues Rule Studio File. Net Content Management Content Repository Rule Team Server Rule Repository Deploy Rule Execution Server Multi Platform Support: Windows/Unix/Linux Server or Mainframe 36 • Java/J 2 EE • z/OS • . NET