Gartner predicted that “Decision Intelligence” will become the dominant trend in 2023 and more than 33 percent of organizations will be using Decision Intelligence.
We talked before about what is Decision Intelligence, and Gartner defines it as:
Decision Intelligence (DI) is a practical discipline used to improve decision-making by explicitly understanding and engineering how decisions are made and outcomes are evaluated, managed, and improved via feedback. The purpose of DI is to make explicit what happens before, during, and after an AI- or human-driven combined decision is made.
Decision Intelligence is taking off, but there is barely any guide on how to start your first project with Decision Intelligence. In this post, we will look at different aspects of it and get you started on the journey of Decision Intelligence with some practical and step-by-step guides.
Step 1: Frame Business Decision
As you have noticed, the name starts with “Decision”. Therefore I’m going to say it loud and clear; decision intelligence is all about “Decision” and let’s be more precise. It is all about Business Decisions.
You might ask, so was the Decision Management, what’s the difference?
Decision Management Suite (DMS) is the evolution of the Business Rules Management System (BRMS). It has all the components of the Business Rules Management System, plus the ability to integrate some form of AI/ML.
The “Decision Management” carries a lot about rules by its natural progression from rules management. As a matter of fact, almost all of the vendors in the “Decision Management” space came from the business rules area (there are some exceptions). It was just renaming (re-branding) the “ruleset” to a “decision” that worked out the job for them.
Though that was fine, “Decision Intelligence” as the successor of the DMS has more to it. As this is the next generation, or we can say DMS+, at its core, must model business decisions.
So the first step in your project is to “frame a business decision” as discussed in this article.
Step 2: The Intelligence
The intelligence part of “Decision Intelligence” is very interesting. That’s because it is about making intelligent decisions, after all. Therefore, you are open to any techniques and opportunities to make the decisions that you have framed in the first step, “intelligent”!
Now you have a clear idea about the business decision that your project is about (because you have framed the business decision in the first step). Therefore, it is becoming very clear which part of the model (the smaller decision units in your model) should use the specific technology for decision-making.
A decision unit in your model might be any of:
- Machine Learning model
- Rule driven
- Computational calculation
- Data aggregation and filtering
- Statistical model and algorithm
- Optimization such as leaner programming, constraint programming, etc.
- … and many other techniques …
But the point is, framing the business decisions allows you to understand the nature of each decision unit and decide on the right technology for it.
Once each unit is satisfied with its decision logic using different technology, the decision graph will bring them together. As you can see here that business decisions by nature are complex, therefore, you will need to integrate multiple technologies to successfully automate and manage them. Or alternatively, an open, unified platform with an integrated approach that should cover your decision intelligence requirements from the technology standpoint.
Keep in mind that the decision graph that brings all decision units together is not just a visual representation but also an executable decision model that integrates all decision units into a unified model – the Decision Graph.
Step 3: Orchestrate
Remember when you are framing business decisions, you model the business decisions with the knowledge of the systems, processes, and information that it requires as input. That’s why your decision graph model specifies the data input dependencies for the related nodes.
Remember, when you frame business decisions, the data inputs are not connected to any data source at that stage to retrieve information from the environment yet.
Connecting to data sources and shaping data happens in an orchestration model. In the model, your orchestration logic is responsible for:
- Collecting the required information from the data sources i.e. database, systems, web services, etc.
- Transforming them into shapes and forms that the decision model (decision graph) expects.
This process is called context building for decision execution. This is a very important stage of decision automation. Many tools and platforms do not have the capability to allow you to build the context in their platform; therefore, it becomes the responsibility of the decision consumer to pass the right information in.
This is a red flag as often data that is needed for decision-making:
- is not exposed to consumer
- requires internal dependencies to other parts of the organization
- is big in size, and a consumer should not retrieve and pass it to the decision
- is evolving and needs the consumer to change, which causes dependency of the consumer to decision internals
Make sure you encapsulate the whole data context creation inside the decision service so the consumer does not need to know how to create the context. So at this stage, the data is collected and transformed, the context is ready, and then the orchestration model calls to the decision for execution by passing the context.
Step 4: Finale
Once the decision is executed, the result of the decisions i.e. the decisions’ outcomes should not just be left unactioned. Decision Intelligence must provide muscle that enables you to integrate the outcome of decisions into organizations’ processes, systems, and running operations.
This step is very important as if you do not integrate the decision outcome into the business operation, you cannot expect any organizational behavior change.
Sometimes this integration of decisions outcome will require dealing with legacy portals without any API or any sort of data interaction layer. In these cases, you can either
- Use a human workflow to assign tasks and delegate the outcome integration to domain experts and other people in the organization
- Automate the outcome integration with Decision Robotics. Meaning a robot can navigate legacy portals and systems and complete the outcome integration without human involvement.
We discussed how to start your first decision intelligence project in this post. Following these 4 simple steps ensures you can utilize the right technology for specific decision units as part of your decision intelligence platform from beginning to end.
One of the important aspects of the efficient Decision Intelligence platform is its ability to bring every decisioning technology under one umbrella for a wide range of leaders; Businesses, Operations, and IT to empower them to improve the speed and quality of key business decisions in changing environments.
As you can see here, “Decision Intelligence” empowers you to think about business decisions in a practical way. This will force you to look at business decisions objectively. Understand them from the context of business. Then pick the right technology to build, automate and execute business decisions. And finally, integrate the decisions’ outcomes back into the organization's business operation.
Last updated November 24th, 2022 at 03:42 pm, Published November 21st, 2022 at 03:42 pm
CEO and the founder of FlexRule – He is an expert in architecture, design, and implementation of operational decisions, business rules, and process automation. Created Decision-Centric Approach, a methodology that brings People, Data, Rules, and Processes together to automate operational business decisions.
This approach is recognized by Gartner as the missing link to provide business value to organizations.