The purpose of this blog post is to succinctly introduce decision management systems for those who are new to the field. First, I present what decision management is, then the technologies that make it up. I also present where they are used and what they bring to the companies that use them.
What is decision management?
There are two types of decision-making technologies. The first are descriptive in that they implement how people make choices among alternatives based on their beliefs and preferences. Think of a doctor deciding a treatment following a diagnosis or a trader buying an asset following a predictive model. The second are normative in that they implement regulations, policies, or strategies regardless of the beliefs or preferences of those who follow the decisions. Think of a loan officer deciding based on an applicant’s repayment history or an insurer calculating the premium an applicant should pay based on the applicant’s medical condition.
There is no single definition that differentiates the two technologies. You have heard of knowledge-based systems, expert systems or reinforcement learning for the former technologies and decision tables, decision trees or business rules for the latter. But there is a consensus to name the second type of technologies decision management systems. So, when you hear or read someone referring to decision management, think of prescriptive methods, technologies, products, and systems implementing formal laws, industry regulations, company policies, and business strategies.
What are the underlying technologies?
Often decision management systems are confused with business rule engines, but they are more than that. Behind the terminology hide multiple technologies. The simplest are decision tables, trees, and graphs. The most sophisticated combine rules and predictive models. If we take the example of SMARTS, it integrates eight decision engines into the same platform. Depending on the problem at hand, one may choose one or the other, or even combine them in the same set-up.
Also, the users of these modern systems are not IT people anymore but business people instead. So, they come with features that allow non-specialists to use them without IT intervention. With modern decision management systems, IT only takes care of the first installations and configurations, the systems come with everything necessary to ensure the governance and security of the applications developed as well as the ease of integrating them into the corporate IT architecture. If we take the example of SMARTS again, it comes with an easy-to-use graphical authoring interface, pre-deployment rule testing, rule repository with version control and rollback, large-scale simulation, real-time decision performance monitoring, and much more.
Where are decision management systems used most?
Like any technology, these systems are not a one-size-fits-all solution for every decision problem. They are not suitable for long-term or midterm slow decisions that companies make once a year or a quarter. In these cases, optimization technologies are more used. They are not also suitable for cases with uncertainty and where probabilistic technologies such as probabilistic graphical models are more used.
Decision management systems are best suited when there is a substantial number of decisions and calculations that are often nested, often invoked, and likely to change often. Therefore, one must consider using them for the operational and day-to-day decisions that companies make in the thousands and sometimes millions in a single day. We find these cases in banking for credit risk assessment, in insurance for premium calculation and even in retail for product configuration.
Although not dedicated to finance, insurance and healthcare, customers of SMARTS have widely used it for loan origination, risk management, fraud detection and money laundering prevention. These are typically cases where organizations make decisions and calculations thousands and sometimes million times a day and may change based on the market dynamics or global economy, or updates to regulations or business strategy.
What are the key benefits of using decision management systems?
Decision management systems come with two critical benefits. As decisions are explicit, they make it possible to understand and explain the decisions implemented so that one can change them more easily when a new situation requires it. They also reduce the risk of errors and biases met in customer-facing applications, such as recommendation, credit, or insurance systems based solely on machine learning.
Key takeaways
- There are two sets of decision-making technologies. The first is descriptive. The second is prescriptive. Decision management systems are descriptive in that they implement formal laws, industry regulations, company policies, or business strategies.
- Decision management systems are best suited when there are thousands or millions of decisions and calculations that are often nested, often invoked, and likely to change often.
- The power of modern decision management systems lies less in the engines but more in the features surrounding them to enable businesspeople take full control of the decisioning process without heavy IT intervention beyond first installation and default configuration.
- Decision management systems have two key benefits. They ease the integration of changes in regulations, policies, and strategies. They also reduce errors and biases in customer-facing applications.
Where to look for further information?
We write regularly about decision management systems and SMARTS. You can find quite a bit of information on our blog and webinars. You can also download our white paper, request a demo of SMARTS, or try it.