Optimal dynamic agent state assignment
View Patent ↗The present invention is directed toward determining an optimal mix of agents, subject to a grade of service (GOS) constraint, for handling outbound and inbound calls in a blended call center environment. Optimality is achieved through the ability to capture real time data via a real time interface (RTI) for each skill-set.
1. A method of assigning a number of agents in a pool of agents to a preferred state and assigning a number of agents in the pool of agents to another state, where the preferred state is one of an inbound state and an outbound state and the other state is one of the outbound state and the inbound state, the method comprising:
determining a first number of agents for assignment to the preferred state based on an expected call rate;
determining a second number of agents for assignment to the preferred state at a first time based on a first call rate sampled at the first time, and assigning the second number of agents to the preferred state;
determining a third number of agents for assignment to the other state at the first time based on the magnitude between the first number of agents and the second number of agents, and assigning the third number of agents to the other state;
receiving a second call rate sampled at a second time;
determining a fourth number of agents for assignment to the preferred state at the second time based on the received second call rate sampled at the second time; and
changing the number of agents assigned to the preferred state by an amount equal to the magnitude between the second number and the fourth number.
2. The method in accordance with claim 1 wherein the preferred state is the inbound state.
3. The method in accordance with claim 1 wherein the first number of agents for assignment to the preferred state is based on a grade of service specification.
4. The method in accordance with claim 3 wherein the grade of service specification comprises a probability that an inbound call will be in a queue for a time longer than a specified time period.
5. The method in accordance with claim 3 wherein the first number of agents for assignment to the preferred state is based on an average call duration.
6. The method in accordance with claim 1 further comprising changing the number of agents assigned to the other state by an amount equal to the magnitude between the second number and the fourth number.
7. The method in accordance with claim 1 further comprising:
receiving agent activity information; and
if changing of the number of agents assigned to the preferred state indicates a decrease in the number of agents assigned to the preferred state, reassigning a number of idle ones of the number of agents assigned to the preferred state to the other state.
8. The method in accordance with claim 1 further comprising:
receiving a third call rate sampled at a third time;
determining a fifth number of agents for assignment to the preferred state at the third time based on the received third call rate sampled at the third time; and
changing the number of agents assigned to the preferred state by an amount equal to the magnitude between the fourth number and the fifth number.
9. The method in accordance with claim 1 further comprising:
determining a rate of change between the call rate at the first time and the call rate at the second time; and
adaptively altering an update interval for call rate sampling.
10. An agent assignment server comprising:
means for determining a first number of agents for assignment to a preferred state based on an expected call rate;
means for determining a second number of agents for assignment to the preferred state at a first time based on a first call rate sampled at the first time, and assigning the second number of agents to the preferred state;
means for determining a third number of agents for assignment to another state at the first time based on the magnitude between the first number of agents and the second number of agents, and assigning the third number of agents to the other state;
means for receiving a second call rate sampled at a second time;
means for determining a fourth number of agents for assignment to the preferred state at the second time based on the second call rate sampled at the second time; and
means for changing the number of agents assigned to the preferred state by an amount equal to the magnitude between the second number and the fourth number.
11. The agent assignment server in accordance with claim 10 wherein the preferred state is the inbound state.
12. The agent assignment server in accordance with claim 10 wherein the first number of agents for assignment to the preferred state is based on a grade of service specification.
13. The agent assignment server in accordance with claim 2 wherein the grade of service specification comprises a probability that an inbound call will be in a queue for a time longer than a specified time period.
14. The agent assignment server in accordance with claim 10 wherein the first number of agents for assignment to the preferred state is based on an average call duration.
15. The agent assignment server in accordance with claim 10 further comprising means for changing the number of agents assigned to the other state by an amount equal to the magnitude between the second number and the fourth number.
16. The agent assignment server in accordance with claim 10 further comprising:
means for receiving agent activity information; and
if changing of the number of agents assigned to the preferred state indicates a decrease in the number of agents assigned to the preferred state, means for reassigning a number of idle ones of the number of agents assigned to the preferred state to the other state.
17. The agent assignment server in accordance with claim 10 further comprising:
means for receiving a third call rate sampled at a third time;
means for determining a fifth number of agents for assignment to the preferred state at the third time based on the received third call rate sampled at the third time; and
means for changing the number of agents assigned to the preferred state by an amount equal to the magnitude between the fourth number and the fifth number.
18. The agent assignment server in accordance with claim 10 further comprising:
means for determining a rate of change between the call rate at the first time and the call rate at the second time; and
means for adaptively altering an update interval for call rate sampling.
19. A computer readable medium for providing program control to an agent assignment processor, said computer readable medium adapting said processor to be operable to:
determine a first number of agents for assignment to the preferred state based on an expected call rate;
determine a second number of agents for assignment to the preferred state at a first time based on a first call rate sampled at the first time, and assign the second number of agents to the preferred state;
determine a third number of agents for assignment to the other state at the first time based on the magnitude between the first number of agents and the second number of agents, and assign the third number of agents to the other state;
receive a second call rate sampled at a second time;
determine a fourth number of agents for assignment to the preferred state at the second time based on the received second call rate sampled at the second time;
change the number of agents assigned to the preferred state by an amount equal to the magnitude between the second number and the fourth number; and
transmit a message for display on an agent terminal providing notification of a change in state.
20. A method of assigning a number of agents in a pool of agents to a preferred state and assigning a number of agents in the pool of agents to another state, where the preferred state is one of an inbound state and an outbound state and the other state is one of the outbound state and the inbound state, the method comprising:
determining a first number of agents for assignment to the preferred state based at least in part on a first call rate sampled at the first time;
assigning the first number of agents to the preferred state;
determining a second number of agents for assignment to the preferred state at a second time based on a second call rate sampled at the second time, the first time and the second time separated by a predetermined update interval;
reassigning a number of agents based upon the magnitude between the first number of agents and the second number of agents to perform a one of: increasing or decreasing the number of agents assigned to the preferred state;
determining a third number of agents for assignment to the preferred state at a third time based on a third call rate sampled at the third time, the second time and the third time separated by the predetermined update interval; and
reassigning a number of agents based upon the magnitude between the second number of agents and the third number of agents to perform a one of: increasing or decreasing the number of agents assigned to the preferred state.