Mixed Integer Programming for Contact Center Agent Scheduling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional scheduling technologies in contact centers are insufficient to handle the complexities and scale of modern contact centers, failing to optimize the number of employees with the right skills at the right time to meet quality of service and cost objectives.
Innovation Solution
A mixed integer programming model is used to determine optimal scheduling of contact center agents based on activity rules, considering constraints and optimization objectives such as minimizing unassigned agents, understaffing, interrupted sessions, and session openings, while ensuring agents are either unassigned or assigned to one session, adhering to minimum and maximum group sizes, and avoiding unscheduling previously scheduled sessions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If traditional scheduling technologies are used, then implementation is simple, but they cannot handle the complexities and scale of modern contact centers
Solution Approach 1:
The scheduling problem is segmented into multiple activity rules, each representing a distinct activity type with its own constraints and optimization objectives. The system processes each activity rule separately through the mixed integer programming model, breaking down the complex scheduling task into manageable segments that can be optimized independently while maintaining overall system coherence.
2Reliability
If more agents are scheduled to meet service level requirements, then quality of service improves, but staffing costs increase
Solution Approach 1:
The system changes the parameter of agent assignment from binary (assigned/not assigned) to optimized allocation based on multiple objectives. The mixed integer programming model adjusts agent assignments dynamically to meet service level thresholds while minimizing the total number of agents required, transforming the staffing approach from over-provisioning to precise optimization.
Solution Approach 2:
The scheduling system applies different optimization strategies to different activity types and time periods. Each activity rule can have customized constraints and objectives tailored to specific business needs, allowing the system to optimize agent allocation locally for each activity while maintaining overall service quality across the contact center.
3Manufacturing precision
If multiple constraints are applied to ensure proper scheduling, then scheduling accuracy improves, but computational complexity increases
Solution Approach 1:
The system performs preliminary actions by pre-defining activity rules with their associated constraints and optimization objectives before execution. The mixed integer programming model is prepared with all necessary constraints (agent availability, minimum/maximum group sizes, session requirements) in advance, allowing the scheduling algorithm to operate efficiently without encountering unexpected computational complexity during runtime.
Data Source
AI summary
A method of leveraging a heuristic-based approach to multi-objective schedule optimization in contact centers according to an embodiment includes adding, by a computing system, assignable contact center agents to pre-existing scheduled sessions, selecting, by the computing system, a session from a plurality of candidate sessions to open in response to adding the assignable contact center agents to the pre-existing scheduled sessions, opening, by the computing system, the selected session, and assigning, by the computing system, unassigned contact center agents to the opened session, wherein at most one session of the plurality of candidate sessions is open for assignment at a given time.