This product handles also partial, incomplete or invalid objects and keeps track of their state. Item, which is incomplete now, becomes complete after enrichment or in combination with other data fragment and may finalize the overall picture like a last puzzle stone.
The input data is passed through multiple processing steps:
- Importing - separation of readable and unreadable parts
- Merging - combination of input data fragments pieces and merging with existing business objects
- Completion - checking of required and optional attributes
- Enrichment - active collection of missing information from remote sources e.g. Web-Extraction
- Validation - validation of completed objects by applying business rules to them
- Release - checking the conditions for data extraction
- Extraction - extraction of data and their transformation into target structures
- Dissemination - export and delivery of data to the target systems
The product has an embedded rule engine (ACTICO Rules). The particular processing steps do invoke rules which in turn make decisions and perform action. Rules work on objects which are passed to them and do control merging, enrichment, completion, validation and release process. A rule action can be anything from a fine granular assignment statement up to the invocation of a high level business function. The rules are defined in graphical manner by the business users themselves and are deployed at into the system at run time.
The rules are edited and maintained with the ACTICO Rules editor by the business users. Finally, the editor generates a java library containing all rules. This library is deployed into the server at run-time without interruption i.e. hot deployment.
The rule editor has large number of built-in functions and actions. Project specific functions and action have been added in order to simplify the rule definition by the business users. The product is open for implementation of high level functions handling the business objects which can be made available within the rules (e.g. event creation). The organization of the rules is fully free.
The processing steps are grouped into jobs. They run in the background and do not interact with the user! Job execution is triggered by events or at defined time schedule. The processing progress is continuously recorded in an event log. The user evaluates the log and performs only actions not covered by the rules. The complete full validation is performed only once after an initial data load. The system keeps track of each object and performs only incremental processing on new or modified objects. For emergency cases when the external data is not available, data can be entered or modified manually. However the completion and validation rules can never be bypassed.
In order to apply different sets of rules to the same data (what-if scenarios) all business objects within the system are under internal version control. Objects (files, persistent data in the Caché database or mix of both) are grouped in so called "Data-Sets" and saved as a Version. Version contains the snapshot of the data. Multiple versions can be created. Versions can be restored later, i.e. the data is restored into its original location. This allows the business users to develop and tune the business rules. It is also used for archiving of processed data.
The product has a Web-GUI for daily business, administration, control and monitoring. It uses simple role-based access control for user authentication and authorization. User authorization via LDAP server is optional.
The product is controlled and monitored with a Web-GUI. Web Service interface is provided for control and monitoring of the scheduler.
An automatic housekeeping function performs cleanup of unusable data and old events.
