This paper studies the resource management problem in IT services, where service request arrives to resource management group (RMG) and RMG needs to allocate resources to a request for a service based on availability of resources and service requirement start date. We propose an approach to find optimal number of resources in the context of Poisson arrivals, and service times & lead-times are exponentially distributed. We provide an exact mathematical queuing model for optimal number of skilled resources needed based on waiting time cost, idle time cost and revenue from a customer service for a guaranteed service level agreement. We tested the model with real life data for various types of requirements and analytical results show the benefits of the proposed approach.