Initiate Master Data Service A Platform for Master Data Management to Help You Know Your Data and Trust Your Data The Hubs: Effectively Managing Specific Data Domains. 3 The Master Data Engine: Processing High Volumes of Data Accurately... 4 Deployment Styles: Four Options to Meet Your Evolving Needs... 4 Integration: Connecting Sources and Systems Efficiently... 5 Clients: Examining and Maintaining Your Data through Graphical Interfaces... 5 Language: Understanding Your Data Regardless of Language... 7 Treating Data as a Critical Asset to Achieve Your Goals: An Example... 7 Summary: Delivering Trusted Data Assets... 8
Truly knowing and trusting - your data empowers you to achieve an array of business initiatives. Whether you re trying to improve customer service, lower operational costs, improve compliance or manage risks, the Initiate Master Data Service platform can help. By delivering a complete, highly accurate, real-time view of data spread across multiple systems or databases, Initiate Master Data Service helps you strategically leverage and share your critical data assets to meet your business s evolving needs. Initiate Master Data Service is a comprehensive platform that enables enterprise-wide master data management (MDM) with a rapid implementation time. The Initiate Master Data Service platform readily evolves as your organization s MDM requirements increase in scope and complexity. The platform delivers trusted versions of critical data assets to users, processes and systems to drive strategic and tactical decision making. It provides comprehensive views of data domains to both end users and automated business processes that rely on these views to function efficiently. Initiate Master Data Service offers several strengths, including: Hubs, specific to a data domain, deliver statistical algorithms, pre-defined metadata and data templates The Master Data Engine delivers high volume matching and linking through high performance data processing and scalable database structures Integration, from the consumption of data from source systems to the dissemination of rationalized data to the enterprise, is achieved through leading-edge middleware technologies. Industry standards and industry-specific requirements are all fully supported. Client tools are end-user applications enabling data stewardship, enterprise search, platform configuration and performance monitoring Finally, Initiate Master Data Service offers a variety of configuration options and deployment styles, industry-standard integration options, graphical-interface tools for end users, and the ability to manage different languages. Read on to learn how the accuracy, scalability, performance, ease of implementation and architectural flexibility of Initiate can help you meet and surpass your business initiatives through better understanding of your customer data. Figure 1: Functional diagram of the Initiate Master Data Service platform Data Sources Sales CRM / CIF Clients Intuitive, graphical user apps Data stewardship apps Implementation & tuning TM Master Data Service Platform Hubs Metadata Data templates Domain-specific algorithms Master Data Engine High-performance execution High-volume data processing High-accuracy match and link Deployment Styles Scoring Registry Hybrid Persistent Languages Localespecific algorithm tuning Crosslanguage matching Consuming Applications Call Center Sales Web DB Integration Leading-edge technologies (SOA) Industry standards (.net, Java, C++, WSDL) Industry-specific interoperability Self-Service Data Warehouse www.initiatesystems.com 2
The Hubs: Effectively Managing Specific Data Domains Hubs encapsulate the experience and intellectual property of Initiate Systems with regard to specific data and data relationship domains, such as consumers, organizations, locations, patients, vehicles, households and hierarchies. Hubs include metadata and instance data. Hub metadata encapsulates the definition of each data or data relationship domain. In addition, the metadata provides the definitions of the high speed matching and linking algorithms that automatically create and manage relationships between data instances. It also defines the composite views, security rules and data quality-error definitions specific to each Hub type. Hubs support the ability to customize metadata definitions to support customer requirements. When implemented in production customer environments, Hubs provide the physical storage for master data and master data relationship instances. For example, the Consumer Hub stores the instance data relating to a business-to-consumer company s consumer customers. Hubs also hold the data quality errors identified by Hub algorithms and record the details of manual and automated error resolution instances. The following data domains have been mastered in production within Initiate Master Data Service : Patient* Provider* Organization* Consumer* Citizen* Location Vehicle Suspect Child Incident Hubs Relationship Hubs Patient @ Provider Consumer @ Provider Consumer @ Organization Hierarchy Household Relationship Hubs capture relationships between and within data domains. For example, Patient @ Provider provides for the identification and management of relationships between patients and providers. The Hierarchy Hub identifies and manages relationships between organizations, and the Household Hub identifies and manages relationships between patients, consumers or citizens. Each Hub can be tuned and configured to your organization s data to optimize matching performance. Hubs contain the information required to execute intelligent parsing of addresses, dates and other attributes specific to the mastered data domain. Fully configurable thresholds can be used to maximize automation of data rationalization. Per your organization s criteria, one or two thresholds can be implemented that categorize processed data within Hubs. In the implementation of two thresholds, for example, processed records are categorized into automatically match and link (above the upper threshold), do not match and link (below the lower threshold), and manual review (between the two thresholds). *Detailed datasheets are available in the Resource Library of www.initiatesystems.com Initiate Master Data Service is a platform that enables implementation and configuration of custom Hubs. For data domains not already mastered, custom Hubs can be developed and implemented within the platform. If a data domain can be modeled in terms of an object with attributes, it can be mastered. The framework of the Initiate Master Data Service platform enables it to co-exist within your current system architecture. Enterprise architecture and data quality are not impacted as your organization grows and your requirements evolve. www.initiatesystems.com 3
The Master Data Engine: Processing High Volumes of Data Accurately The Master Data Engine is the run-time engine that executes algorithms and carries out instructions found in the Hubs. If the Hubs are the brains of the Initiate Master Data Service platform, then the Master Data Engine is its heart. The Master Data Engine is responsible for the actual processing of data that moves into, through, and out of the Initiate Master Data Service framework. The Master Data Engine combines high-volume, high-performance database structures with highly accurate matching and linking technology. The result is a processing engine that out-scales and outperforms competitors in the tasks of acquiring, rationalizing, and de-duplicating data assets. Scaling is accomplished by flattening the data into denormalized data structures. Accuracy rates are higher than competitors because the engine collates a larger set of possible matches and uses value-based matching algorithms instead of deterministic rules which are unable to handle exceptional cases. The de-coupling of the Master Data Engine from the Hubs is a key architectural feature, enabling multiple Hubs to co-exist within a single platform installation, with minimal impact to the overall implementation. Thus, multiple Hubs can be easily added as your MDM strategy evolves. For example, a company can implement Initiate Master Data Service with a Customer Hub and, as requirements change, readily implement additional Hubs such as Organization and the Relationship Hubs of Household and Hierarchy to meet more complex requirements. This architecture also enables the Initiate Master Data Service platform to accept datasets external to the company. For example, a company can receive data from a data service provider such as Dun & Bradstreet and readily consume that data, providing the company with high-confidence, rationalized data that has been enhanced with additional content. Deployment Styles: Four Options to Meet Your Evolving Needs Even though deployment styles are not physical components of the Initiate Master Data Service platform, the explanation of deployment styles lends insight into the architecture of the Initiate Master Data Service platform. Deployment styles represent the variety of methods of storing and federating data throughout the enterprise and making that data accessible by consuming applications. The Initiate Master Data Service platform supports each of the four styles: The scoring style is completely non-invasive to source systems. This style facilitates data exchanges between closed or highly sensitive organizations such as some government agencies. In this style, the Hub delivers only the results of its record analysis and does not retain sensitive information in the Hub itself. In the registry style, a range of data attributes may be retained in the Hub, but data is managed through source systems. Some industries will recognize this model as the classic EMPI (Enterprise Master Person Index) implementation. In this deployment, the record of truth is a virtual record delivered by the Hub. This style works well for organizations seeking to minimally impact their source systems and attain quick time-to-value. The transactional (or centralized or persistent or mastered) style writes and maintains a golden record in the Hub. The Hub retains the master data records, and it is the system of record. Enterprise IDs are propagated to source systems. The hybrid style combines the registry style with the transactional style and enables your organization to be flexible in maintaining and viewing your data. This style allows custom configuration of which systems, source or Hub are the masters at the object and attribute level. Which deployment style is best for your organization depends on several considerations, including end-user activities, audit requirements, speed versus storage issues, future growth plans and constraints on integrating source systems. The Initiate Master Data Service platform provides www.initiatesystems.com 4
the maximum amount of flexibility in deployment styles. Companies can start with one style and migrate to another style over time as needs evolve. Starting with registry style enables companies to achieve value quickly since data does not have to be consolidated in one location. Integration: Connecting Sources and Systems Efficiently The integration components of the Initiate Master Data Service platform facilitate the connection of source and consuming systems to the platform and, when needed, enable data federation throughout the enterprise. Connections to the platform occur primarily through APIs, although other methods of integration are also available. Leading-edge architectures such as SOA (Service- Oriented Architecture) are enabled by Initiate s SDK, and deployment of these architectures is accelerated via Identity Services Publisher. Industry-standard integration technologies such as.net, Java, WSDL and C++ are also supported by the Full SDK. Initiate recognizes that some industries have specific integration standards. These requirements are met through specialized product components, including interoperability options for health care such as EID Synchronization, Enterprise Integrator, HL7 Query Adapter and Message Broker Suite. For high-speed, high-volume extracts from the Hubs to data warehouses or business intelligence applications, the Master Data Extract component is the solution. The Data Federation component allows the platform to provide virtual records composed of attributes throughout the enterprise, including attributes not provided to the Hubs. The Search SDK connects search-only systems to the platform. Clients: Examining and Maintaining Your Data through Graphical Interfaces The Initiate Master Data Service platform includes several GUI clients to examine and maintain your organization s master data. Inspector is a web-based, integrated data stewardship and governance application that includes three components: Inspector for Data Resolution, Inspector for Relationships and Inspector for Data Management. Inspector for Data Resolution: Though the Master Data Engine matches and links at high rates of accuracy, some potential data matches will fall into the manual review zone between lower and upper thresholds. Data stewards resolve and maintain the records that fall into this manual review zone using Inspector for Data Resolution, as shown in Figure 2. In this example, records from several different source systems for Patricia Countryman are displayed, along with the name of the source and the source system s identifier. Inspector for Relationships exposes and maintains relationships between individuals or organizations or other entities that have been declared (perhaps internally or through identification by a third-party trusted source such as Dun & Bradstreet) or derived through Relationship Hubs. These relationships may or may not have been apparent to end users before, but Inspector for Relationships makes these associations between records visually evident, as shown in the screenshot in Figure 3. In this example, relationships are shown between Patty Countryman and other individuals and households in the enterprise. Inspector for Data Management is for organizations that choose a transactional deployment style that persists data in the Hub as described above. Data stewards use this tool (not shown) when they need to add records directly into a Hub independent of source systems. The Workbench tool leverages the Eclipse Integrated Development Environment to configure, manage and deploy Hub metadata and algorithms. Algorithms can be tuned for performance and improved www.initiatesystems.com 5
matching accuracy based on analytical capabilities in Workbench. Rather than have the algorithms displayed in complex code, Workbench (Figure 4) depicts the algorithms graphically. Workbench also supplies user management and configuration management tools for Initiate Master Data Service. The fifth client tool, Enterprise Viewer (not shown), provides a lightweight browser interface for examining rationalized data within the Hubs from anywhere inside the organization. Figure 2: Inspector for Data Resolution displays records from different source systems along with the source s name and identifier Figure 3: Inspector for Relationships displays connections that may not have been evident previously www.initiatesystems.com 6
Figure 4: Workbench enables simplified configuration, management and deployment of Hub metadata and algorithms Language: Understanding Your Data Regardless of Language While the Initiate Master Data Service user interface is available in both English and French Canadian, more important is the ability to understand and process data in multiple languages. To that effect, Initiate Master Data Service operates in all Unicode languages. In addition, Initiate supplies a number of language packs which further enhance data rationalization by providing locale-specific standardization routines that tune algorithms to recognize nuances of locale-specific language, such as name phonetics, name structures and address structures. As an example, a French language pack recognizes that Rue (French for Road ) is statistically insignificant in addresses in French data sources. Language packs are available in the following languages: English, French, German, Spanish, Portuguese, Japanese, Chinese, Korean and Arabic. The Initiate Master Data Service platform also supports cross-language matching and linking. For languages with non-roman characters embedded in them, transliteration is performed, during which each letter in the data is transliterated to its English phonetic equivalent, against which algorithms are run. The Initiate Master Data Service platform currently supports cross-language matching and linking for Arabic and plans to support Japanese, Chinese and Korean. Treating Data as a Critical Asset to Achieve Your Goals: An Example Let s take a look at how a hypothetical consumer goods company would use the Initiate Master Data Service platform to provide better customer service, target its customers more effectively for new products, better assign customers to sales territories and resolve price group membership, and clean up billing issues. The company uses the Consumer, Household, Organization and Hierarchy Hubs, which contain the algorithms, metadata and data templates to understand and rationalize the data domains about its consumer and corporate customers. For the unusual dataset of Sweepstakes, a custom data hub is built to master this data domain that is unique to the company. Implementers familiar with Initiate Master Data Service work with subject matter experts from the company to build this custom data hub using Workbench. Source systems are integrated with Initiate Master Data Service through Java API calls found in Initiate s SDK. The company chooses a registry style of deployment, since it wants to minimize impact to source systems and since the company does not need to create a physical system of record. www.initiatesystems.com 7
Because of the performance and scalability of the Master Data Engine, the millions of records in the various source systems can be processed in sub-second times. The powerful matching and linking capabilities enable the company to automate the majority of error-correction tasks during the early stages of use, even with a low quality of source data. Data stewards resolve any data that falls into the manual review zone by using Inspector for Data Resolution. As data quality improves, Workbench is used to tune the match thresholds, further reducing the volume of records requiring manual review and enabling the Master Data Engine to auto-link at even higher rates. Clean data is provided back to the end users through the various CRM, ERP and other business applications that use this data, and Master Data Extract is used to export this data to the company s BI applications. Later, the company partners with a data cleansing vendor that provides postal service address standardization. After the data passes through Initiate Master Data Service, the company now has confidence that the data sent to this vendor is meaningful and more accurate. The data returned from the vendor is reincorporated into the company s dataset via Initiate Master Data Service. Then, as part of the company s global integration strategy, a subsidiary of the company in Japan has its customer data integrated. The Japanese language pack is used so that the Master Data Engine can transliterate the Japanese data and match and link across languages. At a tactical level, higher quality data reduces costs of communicating with the customer, increases customer loyalty and produces higher revenues per customer. Sales commissions may be reduced through better alignment of customers to territories, and the company can optimize pricing by better understanding who is eligible for a specific price. Though the existing ERP and CRM systems have not changed at all, the company realizes a much higher return on investment for these applications now that the data being used by them is more accurate and therefore of higher confidence. At a strategic level, data becomes information which allows the company leaders to make strategic decisions concerning the company s direction. As the marketplace tightens and makes it more difficult to differentiate among offerings, this company, by treating its data as a critical corporate asset, gains a competitive advantage over other organizations. United States Chicago +1 312 759 5030 Initiate Systems - Austin +1 512 634 5111 Initiate Systems Government Operations +1 703 904 4344 Initiate Systems - New York +1 646 673 8551 Summary: Delivering Trusted Data Assets Initiate Master Data Service is a master data management platform that delivers trusted versions of critical data assets to users, processes and applications. Our data stewardship tools and customerproven approach to matching data help you quickly achieve the business value you need within months, not years. With Initiate s decade-plus of experience in solving data integration problems, we understand that the right version of the truth may differ based on the functional department asking the question. Initiate Master Data Service delivers a trusted version of the truth to the right person, system, or process at the right time. It does so for over 100 production customers across a number of market segments. Why not let Initiate do the same for you? Asia Pacific Initiate Systems Australia Pty. Ltd. +61 (0) 2 8061 3800 Canada Initiate Systems Canada Inc. +1 416 213 8999 Europe, Middle East and Africa Initiate Systems UK Ltd. +44 (0) 118 925 3322 About Initiate Systems Initiate Systems, Inc. enables organizations to strategically leverage and share critical data assets. Our master data management software and experience as an information exchange leader provide organizations with complete, accurate and real-time views of data spread across multiple systems or databases, even outside the firewall. This allows companies to unlock the value of their data assets for competitive advantages or operational improvements. Initiate Systems operates globally through subsidiaries, with corporate headquarters in Chicago and offices across the U.S., as well as in Toronto, London and Sydney. Initiate and the Initiate logo are registered trademarks in the United States and certain foreign jurisdictions. Initiate Master Data Service, Initiate Consumer, Initiate Organization, Initiate Citizen, Initiate Patient, Initiate Provider and Initiate Identity www.initiatesystems.com 8 Hub are trademarks of Initiate Systems. 2007 Initiate Systems, Inc. MDSERV-1107 8