Copyright Â© 2008 James Taylor
. Visit the original article at First Look - Be Informed
Last week I got a chance to catch up with a Dutch company in the decision management space - Be Informed. Be Informed arose out of work within a big systems integrator building complex processes, especially in government, that was unable to find good tools for case management and complex knowledge-based processes. The company started to build a product in 2002-2003 and launched a separate company in 2006 (selling the SI business to Ordina, one of the largest SI’s in the Netherlands). The company has about 130 people split more or less evenly between development and supporting customer implementations. They have seen a steady increase in the scope of projects on which they work and have expanded beyond the Dutch government (although this still represents 50% of their work) to financial services companies in the Netherlands as well as companies elsewhere in the European Union. Historically lots of case management projects but increasingly they are working on more operational projects like deciding which ships to inspect or about awarding specific permits.
They like to talk about separating “the know from the flow” - keeping decisions and processes linked but ensuring that one is not embedded in the other. They find this allows them to define standard process/flow with separate knowledge services that are custom for different customers and to provide process variations. The knowledge services turn a reference process into a “Process of me”. In one agency they took hundreds of processes and reduced them to one generic one using knowledge services that were different, a nice illustration of how decision management can simplify processes.
They believe that if knowledge differentiates you then you should put it at the center. They have what they call a “Knowledge Infrastructure” to develop, record, validate, apply, manage, govern knowledge (rules or policies) and support this by generating forms, case management processes etc. They offer a range of what they call “knowledge instruments” including query tree, classification, navigation, calculator, comparator and search engine. All these instruments use the semantic model for logic, classification, calculation, searching, and so forth.
The product suite consists of:
- ?Be Informed Knowledge Server
Delivers decision services - answers questions for other services and components.
- ?Be Informed Case Management Server
Their process execution environment, focused on case management processes. They also have customers using Siebel, Global 360 and FileNet to execute processes and access the knowledge server.
- Be Informed Studio
Graphic environment for developing the semantic models at the heart of the product plus analysis tools. These models can be executed by the Knowledge Server and the Case Management Server of can be exported to third party tools. One of their clients is working with them to use their product for knowledge management and then exporting the rules to Blaze Advisor for production processing.
- Be Informed Editing Server
Supports editing and collaboration as well as “time travel” (being able to see the state of the repository at a particular time in the past), versioning etc. Models can be compared, linked and impact analysis conducted.
- Be Informed Repository.
At the heart of the product is an Ontology mapped to the business objects that the organization has in its systems and into which the rules are integrated. The knowledge model connects business rules to the ontology while the semantics of the system map to objects in the physical data layer. They use a typical structured natural language approach for rules along with some graphical metaphors. Functions for business rules management, search, information access, governance, case management, dynamic software components and “Knowledge as a service” are all wrapped around this single ontology or semantic model.
One of the interesting side effects of their knowledge focus is that, although knowledge is turned into executable rules for automated decisions, if the system cannot make a decision from the rules then it can present the guiding principles of the knowledge model etc to decision maker to support the manual process - an interesting mix of automated decisions and decision support.