Decision Management 101 Part 3
In the previous post of our Decision Management 101 series, we discussed how decision management can address business growth challenges. If you’re convinced that decision management is the right next step for your organization, what comes next is the age old technology question: build or buy? Like with all technologies, it depends on what time, money, and resources you have available. In order to help you determine whether or not you have the capability to build this yourself, in our last post of the series, we are going to cover what to look for in a decision management system (DMS).
As we mentioned in our first post of the series, decision management gives the subject matter experts and business analysts that are responsible for business outcomes more control over the management of operational decisions. However, whether or not these employees adopt a DMS depends on how intuitive and user-friendly the interface is. Users should be able to spend majority of their time using the tools instead of learning how to use them.
Support for both explicit and implicit knowledge in various formats
Different decisions call for different kinds of knowledge and a robust DMS should be able to support them. For example, our SMARTS™ DMS allows users to author decision logic in various business rule formats (decision tables, trees, graphs, point-and-click rules, syntax rules) and create scorecard, lookup, and predictive models (rulesets, trees, or forests). In addition, users can manage and deploy externally-created PMML and ONNX models.
How well a DMS integrates into your existing systems and processes will also determine whether or not anything gets deployed. For example, can the DMS easily access data from your enterprise data catalogs?
Improving accuracy and performance of decisions requires having the ability to measure decision quality during development, testing, and deployment. Important capabilities to look for include regression testing, simulations, and real-time monitoring.
Lifecycle Management Support
Different activities are required during each phase of the decision management lifecycle (define, develop, test, deploy, monitor, improve). A well-designed DMS should provide efficiency and transparency in each of those phases, including access control, versioning, and release management.
Important security capabilities to maintain integrity of your decisions and control who has access to what, when, and where include authentication, access control, security in programmatic interactions, security in storage, and audit trails. Depending on your industry, certain 3rd-party certifications such as HIPAA in healthcare may also be important to you.
As the decisions being managed impact day-to-day operations, ensuring that your application is up and running is crucial.
Scalability refers to both vertically and horizontally. Often organizations build or buy a decision management solution to solve a very specific use case such as loan origination. However, many of our customers choose a DMS like SMARTS™ because of their ability to use the same platform to manage different decisions across different departments in different regions around the world and to establish auto-scaling policies.
Want to learn whether our decision management system, SMARTS™, is a good fit for your organization? Contact us today to request a custom demo.