Developing strategies for CSAs (costumer assistant agents) of a busy telco. These strategies consist of a roster and policies for the CSAs regarding how to handle incoming communication traffic. Strategies are based on analysing the data generated by simulating the system's traffic using given variables.
To run the output analysis one needs to change the filepaths to the location of the .csv files in own directory in sections Data Retrieval: System 1 and Data Retrieval: System 1.
The objective of this project is to develop a simulation environment (done in Java) which can be run with various sets of input and offers an interface for data analysis of collected data. The data analysis (done in Matlab) in turn uncovers strategies which can be used interchangeably if needed. From this, two leading questions are derived:
- Does one strategy lead to better better performance given the same amount of corporate and customer agents for each shift?
- Can we say with confidence that a chosen roster fulfils all performance requirements and has a certain average waiting time for customers?
The project incorporates many techniques learned in the course Mathematical Simulation and is used to tie the gained knowledge together and to gain practice with simulating a complex system. The project also requires meaningful and statistically coherent data analysis on the output of the simulated system in order to reach a conclusion about, for example, whether one system configuration is significantly better in respect to an output variable than another.