Call Centers & Genetic Algorithms

High Level Overview

Client

Developed a solution for an Italian company that provides other companies with qualified virtual assistants to support their customers.

Problem

They were looking for a way to automate the insertion of breaks in their turn allocation matrix, taking into considerations some constraints on the employees contract type (part-time, full-time) and role (team leader, inbound agent).

Solution

A genetic algorithm that takes in input the turn allocation matrix with no breaks allocated and generates a high number of constraints-compliant assignments at each iteration.
The fitness function takes into consideration an understaffing term and an ovestaffing term, giving more importance to samples that minimize such quantities.