Features

Using the call center simulator, a complete call center network can be simulated on the computer and using the integrated optimizer information about the required number of agents can be obtained.

The call center simulator has many interfaces to handle input data recoreded by the telephone system as directly as possible. Since the simulation models are stored in xml format, at any time a good automatable access to the simulation process is possible.

Performance data of the simulator

  • Full multi-core support
  • Split up simulation and statistics viewer to different computers is possible (client server simulation)
  • Run times of 15-30 seconds for statisticaally stable results on call centers with 100,000 fresh calls per day
  • Platform independend available and command-line controlable
  • Installation and operation without admin rights possible

Input data

The following parameters are required for running a simulation:

  • Client specific parameters
    Configurable per client type:

    • Number of fresh calls per day
    • Distribution of the fresh calls over the day
    • Priority of the clients group
    • Distribution of the waiting time tolerance
    • Retry probability (by first / further retrys)
    • Distribution of the retry times
    • Probability for a client to be forwarded after a talk
    • Probability for changing the client type on forwarding (including the rate for the new client types)


  • Call center specific parameters
    Configurable per call center:

    • Priority for getting assigned clients (per client type)
    • Number of agents in the different agents groups
    • Technical free time when assigning clients to agents
    • Minimum waiting time for clients before they are allowed to enter the specific call center


  • Agent specific parameters
    Configurable per agents group:

    • Number anf working times or avialable number of agent half-hour intervals or available workforce in general and client types which arrival distribution is to be modeled by the angents working time distribution
    • Available skills
    • Service and post-processing time distributions

Simulation process

In the call center simulator the behavior of the clients and agents is simulated using event-driven stochastic simulation. All relevant parameters of the system are recorded in its course over time and as average values continuously.

Output of the results

The simulation results are output in the form of tables and graphs. The results can thereby be displayed directly in the simulator or stored as a html report or provided as csv, xls or xlsx spreadsheets for further processing by other programs. In addition, the entire statistical data can be saved as XML file.

All results can be displayed broken down by client or agent type as average values over the day or on the basis of half-hourly intervals. In this way one can directly verifiy, at which locations and at what times bottlenecks occur.

Using user-definable filters after a simulation run the data can be output in exactly the form how it should be published. Filtering of the statistical data on the side of an external application program is thus not necessary in most cases.

Optimization

Using the integrated optimizer allows to automatically determine the required number of agents that is necessary for a specific parameter to reach the desired target. It can be specified here whether the target value is to be achieved for each client group or for specific groups of clients on average over the day or in every half-hour interval, on average across all client groups. Also you may specify in which agents groups the number of agents is to be changed to achive the target.

Interfaces

The model generator can generate based on csv, xls or xlsx workbooks about caller numbers, etc. a xml client arriving stream. It can be defined largely free in the generator how the input tables are structured (names of each column, customer types to be used, etc.)

The available agents workforce has to be given for every half-hour interval and can also be loaded for all agent groups from a csv, xls or xlsx workbooks. From the numbers of agents per half-hour interval the model generator automatically calculates specific shifts to replicate the workforce distribution.

 

Sitemap  Contact  Data Privacy  Imprint
© TU Clausthal 2019