Zoral Platform

January 29, 2017
“If you don’t have the right data, you can’t make the right decisions...”

Understandably, most loan management systems (LMS) focus on managing loans. They were designed and developed prior to the mass selling of financial products on the Internet. So many struggle to meet the needs of an automated, STP, AI driven world. Zoral Platform (ZDE, BDW, ZML) and LMS solve this issue. They combine to form one of the most powerful credit platforms in use globally today.

The Zoral team is one of the largest in Europe focused on artificial intelligence (AI), machine learning (ML), predictive analytics and automated straight through processing (STP). They have extensive experience. This has been gained by working in the global, consumer and SME digital-lending market for many years.

Our team has developed a next generation Loan Management System (LMS). Zoral LMS manages the entire customer, portfolio and products lifecycle. However, as well as the core loan management functions, it also captures and supplies a large range of predictive data.

This is very important, as it has a profound effect on,

  • ongoing predictive models tuning
  • dynamic customer journey
  • personalising products and services
  • optimizsing operational processes, and
  • improving customer experience.

We found that most of our clients’ LMS systems did not capture all the predictive, decisioning data required. Zoral LMS does. It’s also highly scalable, configurable and extensible. It integrates seamlessly with Zoral Decision Engine (ZDE), Behavioral Data Warehouse (ZBW) and Model Library (ZML), to form one of the world’s most advanced lending platforms.

Background

Zoral LMS helps you manage and adapt your products and services. In today’s environment this needs to be done rapidly, without expensive and time consuming development. LMS does this regardless of…

  • type of lending product
  • which country it is covering
  • portfolio size
  • whether you sell online, offline or both.

Zoral LMS is also ideal for helping you transition from off-line to on-line business.

A survey

As well as our experience with our clients across the globe, we carried out a survey of banking/finance companies. We asked them about the limitations of their existing system. Here is a brief summary of the results.

  • “Lack of flexibility for supported lending product types, customer & product lifecycle processes, customer acquisition channels, (web, mobile, USSD, voice/IVR applications, SMS, chat, etc.)
  • Restrictive e-lending portfolio management, reporting and G/L integration functionality
  • Too much manual effort in the customer & product lifecycle, sub-optimal process automation and scalability
  • Consequences of manual LMS processes were often sited, e.g. “possible to select a wrong option” or errors due to numerous data fields
  • Lack of transparency/traceability of decisions
  • LMS is “stove-piped,” lack of feedback to decisioning
  • Lack of support for predictive analytics
  • Difficult to get data when needed
  • LMS decision workflow hard to customize to match e-lending products requirements
  • Enhancing system functionality and changing business process workflows is difficult and time consuming. , e.g. introducing capabilities such as automated, real-time, decisioning or extending a set of available money disbursement options, often proved to be challenging
  • Limitations for online/mobile lending systems where customer volumes can be significant and growing. Performance, scalability and fault tolerance issues
  • lack of STP (Straight Through Processing) support. A number of LMS’s could not be utilized for online/mobile lending systems
  • In-house developed LMS’s often struggle to get timely and necessary changes delivered. On-line business needs rapid reaction and “Agile” iteration and so are constrained
  • Maintenance/development costs are often disproportionate
  • LMS’s are hard to learn and start using efficiently”

In summary, most banks/finance organizations were restricted by their platform and some struggling to offer e-lending products. Growth and profitability were constrained. Time to market and time adapt and innovate was too long. Using traditional, pre e-lending, LMS platforms was not working efficiently.

Zoral LMS was developed to solve these issues. The next section briefly describes how.

Zoral Platform and LMS

A brief summary of Zoral LMS functionality

  • Supports a wide range of e-lending channels, e.g. online mobile application, chat, SMS, USSD, IVR, customer service phone call session, email, in person at a branch office, via e-commerce, by reference, via marketing campaigns, affiliates, etc.
  • Interacts securely, (synchronously and asynchronously) with devices and applications
  • Interacts securely, (synchronously and asynchronously) with internal business, ‘CORE’ processing systems to improve, automate and scale business processes
  • Provides API based, data/event notification integration points to other enterprise software components (such as CRM, existing LMS, G/L, Billing, ERP etc.). These are flexible, scalable, bi-directional, platform-independent
  • Integrates with key software components (such as a Decision Engine, Behavioral Data Warehouse, messaging systems, Big Data enabled reporting and BI systems, etc.)
  • Supports a large range of lending products e.g. unsecured micro-loans and installment loans, secured loans, collateral management, hire purchase (rent to own) loans, agriculture loans, e-commerce loans, pawn broker loans, car loans, SME loans, line of credit, etc.
  • Powerful tools for business users to configure and adapt products
  • Supports additional, add-on products such as loan insurance, refinance and loan extension process, etc.
  • Configurable product lifecycle, including: application processing, underwriting, contract signing, money disbursement, late payments management, loan recovery, collections and closure, loan extension, loan refinance, legal and write-off, referral, loyalty programs management, promotions, a/b testing, etc.
  • Supports intelligent, AI based, straight-though-processing (STP)
  • Minimizes or eliminates the need for a human interaction
  • Definable, executable, workflows for loan management lifecycle, business processes (e.g. loan application and underwriting, loan recovery and collection, legal, write off etc.). Processes can be fully automated or include manual activities and procedures, (when and as required by the business or regulation/compliance)
  • Designed to utilize synergies between online and offline versions of lending products
  • Configurable to scale to millions of customer/application/device interactions per day
  • Low-latency system response time. Latency is a configurable/granular parameter
  • Decisioning transparency, tracking historic decisioning and LMS initiated changes, (create, update or “soft delete”) made to a customer or a loan record
  • Interacts with 3rd party and internal systems data providers. Re-usable, intelligent 3rd party data/application adapters
  • Extensible, customizable integration with electronic document storage, payment systems, money transfer systems, 3rd party offline services, (e.g. collection firms), and 3rd party online services (e.g. electronic signature)
  • Extensible, customizable data integration via export, or though messaging system, to external systems such as external BI or General Ledger
  • Extensible, customizable General Ledger interface
  • Bi-directional interaction with Decision Engine to support “intelligent” decisions and verification throughout the customer lifecycle using heterogeneous data sources e.g. past and current loan performance, credit bureaux, fraud databases, bank statements etc.
  • Supports a wide range of data including behavioral, social, device, MNO, web, unstructured, ad-networks, application, multi-bureau, marketing, chat, telematics, geolocation, machine, time series, voice, image, biometric, utilities, government, weather data (etc.). “Big Data” capability
  • Configurable to adjust business processes based on changing government regulation worldwide
  • Handles multiple lending products/services and lending processes/workflows across geographies, organization and product hierarchies and other business dimensions
  • Auto-generation and design options, e.g. UI screens or other customer journey interaction methods Includes the creation of and update of mobile device customer application processing, targeted campaigns management, retention management, collections, etc.
  • Integration with popular content management systems (such as WordPress or Drupal) to integrate processes managed by the LMS into existing websites and mobile applications
  • Creation of dynamic customer journey across different channels. Mass personalize lending products and services
  • User definable lending product attributes e.g. interest rate, fees, current balance or credit limit calculation formulae, refinancing options across geographies and regulatory regimes worldwide
  • Powerful, easy-to-use, version control for work flows and rules. Captures business user know-how and business intelligence, (e.g. setup and maintenance of calculation of interest and fees for a loan, for a specific product or A/B testing of a product idea). Comprehensive business process work flows (e.g. loan application, underwriting, retention, cross-sell, upsell and collections);
  • Powerful rules engine to enforce business and compliance policies such as credit amount or credit limit terms, on customer and portfolio level
  • Re-usable rules and work flows, as building blocks of automated e-lending processes management
  • 360° customer view, consolidating data from multiple data sources including,
    • customer channels
    • current and historical information
    • customer lifetime value (LTV)
    • associated costs (e.g. marketing and operating)
    • past and current loan status and history (including arrears)
    • history of payments (including late payments)
    • risk- and marketing-related scores and indicators (on customer and products/loan level)
    • credit bureau reports
    • customer-related warnings and notifications
    • related customers (e.g. referrals, family members or friends).
    This data is utilized for each validation and decisioning step as part of customer lifecycle management process.
  • Configurable to support
    • funding management
    • cash management
    • marketing (e.g. loyalty programs, up-sell, cross-sell)
    • compliance management
    • time management (e.g. calendars, promises to pay)
    • asset management (inventories)
    • collateral management
    • 3rd party management (commissions, asset disposition, broker, etc.)
  • Supports online/mobile customer self-service functionality to reduce operating costs
  • Settings synchronization with existing enterprise systems
  • Built in data quality indicators and scores
  • Automated tracking of customer communications history e.g. phone, SMS, email, chat communications. Integration with telecommunication, email and instant messaging systems
  • Defines new loan lifecycle management processes from scratch or encapsulate sets of rules and work flows as intelligent, version controlled templates. These can be re-used/modified across products, geographies and other business dimensions
  • Monitoring of current and historical KPI’s and metrics at product and portfolio level;
  • Lending process workflow testing functions, including unit testing, batch testing, automated/regression testing
  • A/B testing to facilitate product tuning, innovation and optimization of loan application and contract management processes
  • Scalable reporting and BI as a component (operations, financial, marketing, GL) and real-time input source to enterprise, Big Data architecture
  • Granular, auditable security and user access. Integration with existing directory services (LDAP, Active Directory)
  • Fault-tolerant, scalable
  • Supports loan securitization
  • Modular deployment, modular structure i.e. only modules that are required need be deployed and maintained
  • Ability to monitor system and portfolio performance in near real-time
  • 24/7 fault tolerant architecture
  • Supports long running business processes. For example, some activities and procedures of a loan application or contract management process may require human interaction, (duration might be from minutes to days)
  • Separation and control of business workflow development and its approval from UAT environment to production
  • Deployment to public cloud, or hosted on enterprise private cloud or multi-instance deployment to a server farm, or provide support for a hybrid cloud/server deployment.

Given the above, Zoral LMS and platform is ideal for,

  • Intelligent STP enabled, scalable, online/mobile e-lending
  • Intelligent off-line lending
  • Blended online/offline lending strategy
  • E-commerce based lending
  • SME lending
  • MNO based e-lending
  • and many others

Zoral platform and LMS architecture

The diagram below shows the Zoral LMS component high-level architecture:

LMS Architecture

oral LMS is modular, comprising several scalable, fault-tolerant components. They key ones are briefly described below.

LMS - Process Runtime

Process Runtime is the core LMS component, responsible for coordinating and executing customer & product lifecycle management processes

  • Customizable
  • Capable of supporting multiple lending products simultaneously
  • Supports the main concepts of Business Process Management Notation 2.0 (BPMN) providing the ability to automate complex business processes, including processes/sub-processes, activities and tasks (including manual), decisions, events, collaboration (messages), gateways and choreography
  • Lending business processes are defined declaratively and can be created and modified without software development effort. Business process definitions are versioned. Changes are logged (with a granular, time-series audit trail) and can be reverted
  • Process execution meta-data is stored in the Process Runtime Database and can be utilized to build predictive models and mine knowledge for further optimization
  • Process Runtime is the main component communicating with existing internal core processing systems and enterprise software system components , (via platform-independent API’s such as SOAP, RESTful, messaging etc.). It delivers the necessary set of customer & product lifecycle management processes to the enterprise architecture
  • Can be customized to interface to a number of Reporting/BI and General Ledger components.

Domain Model Runtime

The Domain Model Runtime defines and customizes complex domain models for different lending businesses, including creating entities with pre-defined and custom attributes (such as customer, lending product, loan application, loan contract, collateral, etc.). You can also create entities relationships, data validation and data transformation rules without software development effort. Similar to business process definitions, domain model definitions are also versioned. Changes are tracked, logged and can be reverted.

  • Utilized by the Process Runtime component for data management of the lending operations such as creating, modifying and soft-deleting customer data, applications and contracts data and data management of loan transactions. The LMS data management logic is not hard-coded in any part of the business process definition and can be defined and modified using a scripting language (e.g. JavaScript);
  • The Domain Model Runtime supports the concept of business (or logical) transactions, “wrapping” a set of database transactions to define and manage transaction boundaries of a business operation, e.g. money disbursement, which requires successful completion of multiple steps before the entire operation can be considered as complete and committed to the database
  • This component supports different concurrency models including pessimistic and optimistic locks
  • Data produced as result of customer & product lifecycle management processes is stored in the LMS’ Operational Database.

UI / Frontend Runtime

UI Runtime handles the user interface, (layout, look & feel etc.), and presentation logic, (e.g. dynamic user journey), for various frontend components such as customer websites or mobile/USSD/voice applications and front office websites.

  • UI Runtime is optional. It is required when user-facing parts of the solution are delivered using the Zoral LMS rather than by existing frontend systems developed by the business
  • It communicates with the Process Runtime component to deliver a dynamic user journey (e.g. which screen needs to be shown next or which additional UI elements need to be displayed based on decision engine, LMS process workflow and business logic);
  • The UI / Frontend Runtime component supports different kinds of user interface technologies. For example, it supports traditional web UI (pure HTML) and rich web UI (AJAX, AngularJS, jQuery etc.), USSD menus/applications, mobile applications (iOS, Android, Windows Mobile). It can be extended to support other types of user channel interaction.

UI / Frontend Runtime supports,

  • Automatic generation of the UI “on the fly” (in runtime) or when designing a solution for a lending product. This is based on existing domain model and business process definitions. This option is suitable for cases when speed of developing/iterating a user interface is important
  • It can be used to generate quickly a user interface, then fine-tune it manually, applying web design best-practice and customization to suite the brand image of the e-lending business
  • It also allows manually created/altered user interfaces to be connected to the Process Runtime component
  • If utilized and configured, the UI / Frontend Runtime automatically attaches the Zoral BDW for automatic behavior data capture

Another important feature of the UI / Frontend Runtime is that it can be integrated with an existing Content Management System such as Drupal or WordPress) to embed screens and forms of the solution into the existing web sites. This can greatly simplify the maintenance and modification of the e-commerce/lending website. The UI generated by the UI / Frontend Runtime LMS component can also be customized to “inherit” styles from the target website.

Notifications Runtime

This is an optional component responsible for communicating with customers and lending business personnel, by sending notifications and messages. This is managed and requested from the Process Runtime component) via different channels. These include automated phone calls, SMS/MMS, email, IM messages etc. The component supports the receipt of inbound messages and forwarding of messages to the Process Runtime component to track the full history of communications within a lending business process and drive the business process logic based on bi-directional customer messaging.

Configuration & Admin Runtime

This is a key LMS component. It stores and provides all process, entity, data integration and UI definitions and configuration settings defined for the deployment, including:

  • Business processes, domain models and UI definitions, across lending products, businesses, geographies etc.
  • Operating and integration settings for each LMS component
  • Global and local settings of the deployment.

Each definition or setting is versioned. The changes are logged into an audit trail and can be tracked.

Configuration & Admin GUI

This is an optional component providing a rich, web UI to manage definitions and settings for Configuration & Admin.

The components of the Zoral LMS communicate with each other via clearly defined synchronous and asynchronous internal interfaces or APIs.

Each component is horizontally scalable to cope with increasing load and is fault-tolerant, providing uninterrupted and reliable execution of mission-critical processes. LMS’ architecture was designed to be resilient, 24/7 and avoid single points of failure.

To integrate with external systems, LMS components have external interfaces. These are platform independent. The following protocols are supported: SOAP, RESTful, RabbitMQ. Support for additional protocols can be added. The platform-independent interfaces allow seamless integration of the Zoral LMS with existing CORE processing systems and enterprise software systems components, regardless of their platform and technology stack.

Zoral Loan Management System supports multiple deployment models: in a public or private cloud, on-server farms or supports a hybrid server/cloud deployment model.

Conclusion

Zoral LMS is innovative and meta-data driven. Its component based approach enables financial institutions seeking to deploy digital lending products. It overcomes the limitations of traditional Loan Management Systems as outlined above.

Zoral LMS provides a solution that is

  • modular
  • scalable
  • low-latency
  • fault-tolerant,
  • intelligent,
  • automated
  • STP enabled
  • extensible.

It can rapidly be adapted for each digital product, across geographies. It avoids “hardcoding” the solution by providing a way to declaratively design/customize the key components and business processes. Plus, this can be done by business users rather than developers.

With Zoral LMS you have the ability to launch new or enhanced digital products quickly and to iterate based on early feedback. This significantly reduces time to market and enables the fintech model…

rapid, low cost, profitable, adaptable digital product deployment.

This document is provided by Zoral Limited and its affiliated companies (“Zoral”) for informational purposes only, without representation or warranty of any kind. Zoral shall not be liable for errors or omissions with respect the information contained in this document. Product Specifications are subject to change without notice. The only warranties for Zoral products and services are those that are set forth in the express warranty statements in Zoral’s standard contracts for such products and services, if any. Nothing herein should be construed as constituting an additional warranty.

© Zoral Limited 2016