Automated Change Orchestration with Request Subscriptions
Traditional master data management enforces rigour and discipline in how changes are orchestrated. Most end users, however, care deeply about specific applications and would much rather that changes were synchronized without elaborate policy definition. Oracle Enterprise Data Management Cloud takes a crowd sourcing-style approach that allows one to ripple changes made in one context across alternate perspectives or entirely distinct applications. This is accomplished by defining request subscriptions.
Whenever a request is successfully applied on a source dimension, emails notify subscribers of a subscription request that conveniently prepares the same change in the context of their target business perspective. The assignee may now choose to accept the change, enrich it, and commit to it or delete it entirely. This incremental, agile, informal and organic change process helps changes in one application flow across consuming application contexts ensuring eventually consistent, aligned shared business perspectives.
Inference-Based, Collaborative Governance Workflows
Oracle Enterprise Data Management Cloud (EDMCS) takes a novel approach to data governance that draws upon granular permissions for secure data access, approval policies to identify signatory authorities, and a data-driven approach to ensure awareness of request context. Traditional workflow tools take a declarative approach that requires orchestration of workflow stages, tasks, and related artefacts to model a business process.
Oracle Enterprise Data Management Cloud (EDMCS) uses an intelligent inference engine to combine knowledge of the request content, user permissions, authorized approvers, and request stage to invite the right stakeholders to participate in a workflow, while uninviting others whose participation is no longer required making for a dynamic, self-managed orchestration. Participants can collaborate via conversation threads to discuss issues either at an item or request level.
When user load items from a file, the files are logged as attachments, complete with annotations for items that were either ignored or skipped on load. User attachments may also be included to justify or supplement a request. Request history is also included to audit who made, what change, when and where in a change request.
Application Adapters: Packaged integrations & open Interface
Oracle Enterprise Data Management(EDMCS) Cloud introduces the notion of an application adapter. This feature is foundational to connecting with, interpreting the metadata for, and communicating changes with a connected business application. Two types of adapters are available: custom or tailored. Tailored application adapters are intended to deliver packaged integration, while the custom adapter is intended to integrate with any application, on-premises or in the cloud, Oracle or non-Oracle.
Example tailored application adapters include linking to other business processes in the Oracle EPM Cloud, such as Planning, or to the chart of accounts in Oracle ERP Cloud. Such adapters demonstrate awareness of the connected application, come bundled with specific properties, application metadata and application-specific validations to establish high quality connections with the connected application. Application administrators may use registration wizards to answer a set of questions to establish a connection, ingest application-specific metadata, and instantiate an application-specific view with which one can synchronize data imports and exports.
Separate perspectives called viewpoints are created within each application view to give end users window into each dimension such as account, product, entity, cost center, et al. One may maintain enterprise data elements, their attributes, relationships, associations, usage within related application objects such as plan types, and synchronize changes to enterprise data, e.g., dimensions, data mappings, back to the external application. Connected applications are integrated seamlessly via web services based upon RESTful interfaces. Other applications are integrated via a flat file interface.
Tools of the trade: Request, Validate, Compare
The Request experience embeds data quality directly within the change management experience by enabling real-time system checks, duplicate prevention, control for shared node insertion, and application specific validations. While the request experience allows end users to create change requests either interactively or by loading items from file, tools such as validate and compare tools provide a workbench experience for data stewards to perform analysis and generate new items for remediation purposes. During application onboarding or during on-going change management, stewards can validate entire viewpoints and work through lists of issues.
As data stewards correct errors, request items are recorded, visualized and available for them to commit. Similarly, data stewards may also run comparisons between alternate hierarchies or lists and hierarchies to identify node differences. Differences can be generated across entire structures or for bottom-level records alone.
Users can locate similar records across different viewpoints, and visually inspect any property differences. They can record items for remediation within requests either by editing properties in a viewpoint or by simply dragging and dropping records across viewpoints to rationalize them.
Node type converters transform records as they cross application boundaries to add qualifiers (e.g., pre-fixes), and copy attribution based upon pre-defined property maps across application-specific node types. Several business scenarios are enabled via this conformance and rationalization process, including dimension conformance to power trustworthy analytics, data rationalization to harmonize across applications, and data transformations across transactional and analytical applications. End users may use the comparison process to build data maps as well. Once both dimensions and mappings are prepared, they are ready for consumption.
Consume Application specific changes downstream
The consumption of enterprise data can take many forms. Administrators may configure viewpoints to enable read-only browse experiences for casual business users. Entire viewpoints may be downloaded to an Excel spreadsheet. This can be used, either for offline request generation or to share information within workgroups via email. Changes can be uploaded to the same viewpoint, e.g., mass updates, or be repurposed to share data across viewpoints (e.g., to power plan with production as a starting point). Further, stewards can synchronize registered applications by exporting dimensions and or data maps on-demand.
The same process can be automated using scripts that may be run using the EPM Automate utility. Lastly, read-only access may also be provisioned to auditors so they may browse enterprise data, inspect request histories, and learn about all changes applied to production.