Skill capacity planning and transformation scheduling of IT workforce under stochastic learning and turnover

Most IT service providers rely on myopic hiring strategies to meet immediate demand resulting in both demand gap and agent glut over time. We show that the provider may have an incentive to reject demand, even when it has sufficient skilled service agents to meet the demand, if it perceives value in inducing agents into learning. This would enable the skill supply to align quickly to the changing demand and thus earn future profits at the cost of immediate revenues. We consider a skill capacity planning problem while minimizing a sum of skill demand-supply gap over a finite planning horizon. We develop a discrete time Markov decision process framework and design optimal demand admission control, hiring and skill transformation strategies under stochastic rate of turnover and skill learning. We develop a piece-wise linear demand admission and hiring control rule. We show that there are states where it is optimal to reject a part of the demand and induce agents into skill learning. An efficient computational algorithm is designed to evaluate the control policies.