CareerCup

based on past experience and details available to them. ask managers to estimate on incoming customer tickets rate and ticket resolution rate per resource. keep reviewing the plan once every week.

pre- production issues:
no. of resources needed = [ no. of existing tickets + ( new ticket arrival rate per day - tickets fixed per day) * number of days remaining to close all tickets ] / tickets fix rate per resource.

for post production issues:

number of resources = number incoming ticket * average effort in hours per tickets / service level agreement (in hours)

Could you please provide all the available attributes. Like the shifts, number of calls, region etc.

I would build a system using Decision trees, but that initially requires a Data Set which i believe would be the input.

You could use a supervised algorithm for this.

Provide the number of people required on a given day at a given time for a time-frame (say past one month). A good algorithm can predict the number required now.

Or in a crude manner, base the required number off last year's data, allowing for a buffer, such as lesser employees etc