ST⛔P treating #ai #machinlearning #ml #llms models as business decision.
💢There is a big gap between the outcomes of #ai #ml #llms and what business decisions are. In many organizations, these models or their outcomes are interchangeably used as a synonym for business decisions. Data and analytics functions in organizations try to improve the quality of business decisions without even modeling the business decisions, and their assumption is that the technical model of #MachineLearning #ai #llms is “the” business decision.
💥The intention of building these data-driven and technical models is, for sure, to improve business decisions (at least most of the time) and create better business outcomes. However, there is a big misunderstanding about what business decisions are within the #data and #analytics leaders.
✨Business decisions are business questions seeking answers. They are the very core of business operations and the organization’s value. Focusing on managing and automating them is a no-brainer. But this should not become the base for inferring they are not the #ai #ml #llms #process models that we use to predict, automate, etc. They are separate from the technical models we create in data and analytics practices, and they are the very core entities of any organization that should be managed explicitly.
🤔So now the question becomes how to capture (model) and communicate business decisions?
The best way to do this is using the decision graph, which enables leaders to model complex, hierarchical, and multi-step business decisions. This model clearly lays out how a decision is made. Hence, it is a simple visual model abstracting the technology complexity out; it will communicate well to every stakeholder how decisions are made.
Once the team models the particular business decision, this model becomes the blueprint for all its dependencies. Any business decision depends on required data and information, systems, technology, and even other business decision outcomes (sub-decisions), as they don't live in isolation.
📖Learn more at https://lnkd.in/gUubs6N9 on how to model business decisions.
🚀Once the business decision is decomposed into smaller, easier-to-understand decision units, it means you modeled them explicitly, then you can use any data and analytics techniques such as #ai #machinelearning #rules and etc. for any of those sub-decisions and decision units in the graph. Modeling business decisions explicitly and managing them as separate entities will enable organizations to make data and analytics useful, change organizations' behavior positively, and improve the effectiveness and efficiencies of their operations.
#decisionintelligence
Posted here.
Published April 19th, 2024 at 07:30 am
