✨A decision model with composite AI capability is the core foundation of Decision Intelligence where #llms #genai remain💥hopeless!
The Open Decision Intelligence Platform (DIP) enables organizations to build a decision model that combines multiple AI techniques into one coherent model for decision-making. What is essential is the coherent model that brings together various AI techniques such as #rules #machinelearning #optimizations.
💡Note that this model is very different from a business process as they serve different, distinct purposes. The decision model based on decision graph technology enables coordination between multiple AI techniques.
🤔What does that look like? What is an example scenario?
Imagine a retail store where employee schedules need to cover the peak hours, while the company should minimize the cost of labor and ensure more senior staff are available to support teams.
We can create a decision graph that has multiple nodes combining the decision units below:
🔹Machine Learning: a model based on historical data to predict the sales volume, store traffic, and busy hours.
🔹Business Rules: to determine the boundaries of working hours for staff based on leave availability and other companies' policies and the staff eligibility
🔹Optimization: Create an optimal employee schedule considering the forecasted demand for employees' availability for support teams and senior team members. Ensure each team has the right number of senior and domain experts available while minimizing operation labor costs.
As we see in the above example scenario, multiple techniques of #ai, such as #rules #ml #optimization, are used in a decision model where the outcome of the model is the employee's schedule for peak hours.
🚀Although this is just one example in the field of #DecisionIntelligence Platform #DIP, there are limitless sets of scenarios that operation teams need to take care of across various activities and industries in #insurance #financialservices #banking #government #health #logistic #energy #utilities #retail and more!
Posted here.
Published May 24th, 2024 at 07:30 am
