How to implement explainable AI in scheduling?

To implement explainable AI in scheduling using the Solvice APIs you will use the Explainable AI feature. This feature is designed to offer a deeper understanding of how decisions are made within our solvers.

Here’s how it works:

1. Triggering the Explanation: The process starts when a user requests an explanation for a particular decision. This is because the computational effort required is significant, so the feature is activated on demand.
2. Hyperlocal Discovery Phase: Once triggered, the Explainable AI feature enters a hyperlocal discovery phase. This phase involves a detailed analysis where the solver reevaluates the decision by considering all possible alternative assignments for the questioned decision.
3. Evaluation of Alternatives: The feature then calculates a score for each alternative assignment, providing a comparative measure of how effective each alternative would be compared to the chosen solution.
4. Providing Insights: This score evaluation helps users understand not just the optimal solution but also the quality and impact of potential alternatives. It reveals why certain options were preferred and how different choices could influence the overall effectiveness of the schedule.

This approach helps clarify why specific resources are assigned to certain tasks and allows users to see the potential outcomes if different decisions were made, enhancing transparency and trust in the AI’s decision-making process.

‍