Method, apparatus, and computer-readable medium for managing concurrent communications in a networked call center

Abstract

A method and apparatus for scheduling agents in a call center to meet predefined service levels, wherein communications are associated with queues representing categories of communications, the queues including at least one concurrent queue of concurrent communications, wherein multiple concurrent communications can be handled concurrently by a single agent. The method includes executing a simulation to determine an effectiveness of plural agents. The simulation includes computing a skill group weighting (SGW) for each agent for at least one concurrent queue and at least one interval based on: tc, the time spent by the agent on queue C communicationstall, the time spent by the agent on all concurrent communicationste, the elapsed concurrent time for the agenttn, the non-idle time of the agent; andAgents are scheduled based on the SGW and max capacity of concurrent communications for each agent.

Description

FIELD OF THE INVENTION

The invention relates to managing concurrent communications in a networked call center by scheduling agents for handing the communications in a manner that achieves desired service levels.

BACKGROUND

Assigning workers to shifts in a manner that allows the workers to handle tasks in an efficient manner is a critical part of many businesses. For a business such as a contact center (also referred to herein as a “call center”), workers (e.g., agents) are assigned to tasks (e.g., incoming communications) based on skills associated with each agent and the skills required for the tasks. One mechanism for matching the communications with the skills of an agent is to associate the communications with a “queue” that represents a category of the communication, such as Technical Support, or Billing Issues. Agents with the requisite skills can then be assigned to one or more appropriate queues over specific time intervals.

As may be appreciated, when an agent has the requisite skills to work multiple queues, the call center may have difficulty determining which scheduling assignment is optimal because there is no easy way to see how the agent is contributing across all of their queues. One solution is simulating the work on all of the queues with different agent assignments. However, to handle communications most efficiently, agents with the requisite skills must be scheduled for times when communications requiring those skills are likely to be received. Matching agents to communications while maintaining efficient staff levels and meeting requisite service levels is a highly complex process.

SUMMARY

The disclosed implementations address an agent's contribution to each of the queues taking into consideration the concurrent maximum communication assigned to each agent. One aspect of the invention is a method for scheduling agents in a call center to meet predefined service levels, wherein communications are associated with queues representing categories of communications, the queues including at least one concurrent queue of concurrent communications, wherein multiple concurrent communications can be handled concurrently by a single agent, the method comprising: executing a simulation to determine an effectiveness of plural agents, the simulation including: computing a skill group weighting (SGW) for each agent for at least one concurrent queue and at least one interval based on:

• • t_c, the time spent by the agent on queue C communications
  • t_all, the time spent by the agent on all concurrent communications
  • t_e, the elapsed concurrent time that the agent has spent on communications in the interval
  • t_n, the non-idle time of the agent; and

scheduling the agents based on the SGW and max capacity of concurrent communications for each agent.

Another aspect of the invention is a system for scheduling agents in a call center to meet predefined service levels, wherein communications are associated with queues representing categories of communications, the queues including at least one concurrent queue of concurrent communications, wherein multiple concurrent communications can be handled concurrently by a single agent, the system comprising at least one computer hardware processor and at least one memory device storing instructions which, when executed by the at least one processor, cause the at least one processor to carry out a method of: executing a simulation to determine an effectiveness of plural agents, the simulation including: computing a skill group weighting (SGW) for each agent for at least one concurrent queue and at least one interval based on:

• • t_c, the time spent by the agent on queue C communications
  • t_all, the time spent by the agent on all concurrent communications
  • t_e, the elapsed concurrent time for the agent
  • t_n, the non-idle time of the agent; and
  • scheduling the agents based on the SGW and max capacity of concurrent communications for each agent.

Another aspect of the invention is non-transient computer-readable media having instructions stored thereon which, when executed by a computer processor, cause the computer processor to carry out the method comprising: executing a simulation to determine an effectiveness of plural agents, the simulation including: computing a skill group weighting (SGW) for each agent for at least one concurrent queue and at least one interval based on:

scheduling the agents based on the SGW and max capacity of concurrent communications for each agent.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a schematic diagram of a call center computing architecture in accordance with a disclosed implementation.

FIG. 2 is a schematic diagram of system architecture for incorporating a scheduler into a contact center in accordance with a disclosed implementation.

FIG. 3 is a flow chart of a process for calculating skill group weights in accordance with a disclosed implementation.

FIG. 4 is a flow chart of logic for creating a staffing schedule in accordance with a disclosed implementation.

FIG. 5 is a block diagram of a computing device that can be used in disclosed implementations.

FIG. 6 is a Gantt chart of an example of an agent handling communications in accordance with a disclosed implementation.

FIG. 7 is a Gantt of another example of an agent handling communications in accordance with a disclosed implementation.

FIG. 8 is a flow chart of logic for creating a staffing schedule, taking into consideration concurrent communications, in accordance with a disclosed implementation

DETAILED DESCRIPTION

U.S. patent application Ser. No. 16/668,525, the disclosure of which is incorporated herein, discloses that a simulation can be performed for a contact center with multiple queues. In a simulation, agents from the skill groups are assigned to the queues based on the skills associated with the agents in a skill group (a group of agents with common skills) and the skills required by the queues, and predicted communications are matched to the queues by a matching engine. The simulation may be performed multiple times over multiple intervals. After the simulations are complete, for each skill group and for each interval, the amount of time that each agent spent working using each skill associated with the skill group is determined for each interval, and an average time for each skill is calculated across all of the agents for each interval. U.S. patent application Ser. No. 16/744,397, the disclosure of which is incorporated herein, discloses how to provide simulation for “deferred” communications, i.e., communications for which a substantially immediate response is not required.

The average times for each skill associated with the skill group is used to create a skill group weight for the skill group for each interval. When a scheduling engine is determining which queue to place an agent in for one or more intervals, the skill group weights for the intervals are used to calculate a score for some or all of the queues based on different placements of the agent. The placement that results in the best score (e.g., lowest) may be implemented by the contact center in agent scheduling. Because the simulations are used to generate the skill group weights ahead of time (i.e., before the agents are scheduled), the agents can be quickly and efficiently placed in queues without having to simulate the queues each time a placement is needed.

During a simulation, a call center scheduling algorithm attempts to place shifts in a fashion that maximizes their utility, for example guaranteeing a desired service level while respecting legal constraints and minimizing operational overhead. When scheduling, agents can be allocated to a shift and the impact on the workload and service levels of each queue at each interval of that shift can be assessed. This impact is directly related to the agent's contribution to the work capacity.

Disclosed implementations are discussed in the context of a call center. However, the innovations disclosed herein can be applied to directing any items or tasks to a specific party, such as an agent or other service provider. In a traditional contact center, an agent picks up one communication (like a call) at a time, works on it, then moves on to the next item. Such communications are referred to as “non-concurrent” herein. However, certain communications, such as chat or social media communications, allow an agent to work on multiple communications at the same time. Such communications are referred to as “concurrent communications” herein. For example, an agent can pick up multiple chats and work on all of them before finishing even the first one.

Each agent can handle a maximum number of concurrent communications based on, for example, the agent's experience, skill levels, and mental acuity. Therefore, the maximum number of concurrent communication of an agent is a personal setting, defined by the agent and/or their supervisor, for example. Therefore, an agent that can handle a maximum of 4 concurrent items has the same throughput as 2 agents that can handle a maximum of 2 items each. Further, an agent that is already engaged in a concurrent item can start another concurrent item, but not a non-concurrent item (like a call). Also, it might be desirable to require that an agent can only start a non-concurrent item after finishing all the concurrent ones because, by definition, a concurrent item requires the agent's attention constantly until completed. when routing communications, a concurrent communication can be assumed to take a percentage of the agent's attention, e.g., 20% for some queues and 50% for another. In this case, a communication can only be routed to an agent who is idle or is working on concurrent items and has a percent attention open which is greater than or equal to the percent attention required for the new concurrent communication. In this alternative, the max concurrent items used in the shrinkage calculator is computed as 1/(percent attention requires). For example, a queue that takes 20% of the agent's attention would have a max concurrent of 5.

FIG. 1 illustrates components, functional capabilities and optional modules that may be included in a cloud-based contact center infrastructure solution. Customers 110 interact with a contact center 150 using voice, email, text, and web interfaces to communicate with agent(s) 120 through a network 130 and one or more of text or multimedia channels. The agent(s) 120 may be remote from the contact center 150 and may handle communications with customers 110 on behalf of an enterprise. The agent(s) 120 may utilize devices, such as but not limited to, workstations, desktop computers, laptops, telephones, a mobile smartphone and/or a tablet. Similarly, customers 110 may communicate using a plurality of devices, including but not limited to, a telephone, a mobile smartphone, a tablet, a laptop, a desktop computer, or other. For example, telephone communication may traverse networks such as a public switched telephone networks (PSTN), Voice over Internet Protocol (VoIP) telephony (via the Internet), a Wide Area Network (WAN) or a Large Area Network. The network types are provided by way of example and are not intended to limit types of networks used for communications.

In some implementations, agents 120 may be assigned to one or more queues 125, and the agents 120 assigned to a queue 125 may handle communications that are placed in the queue by the contact center 150. Agents 120 and queues 125 may each be associated with one or more skills. The skills may include language proficiency (e.g., English, Spanish, and Chinese), proficiency with certain software applications (e.g., word-processors and spreadsheets), training level (e.g., having taken a particular course or passed a particular test), seniority (e.g., number of years working as an agent 120), achievements (e.g., meeting certain performance or quality goals, receiving positive performance reviews, or receiving positive reviews or ratings from customers 120). Other types of skills may be supported. The skills associated with an agent 120 may be the skills that the agent 120 possesses. The skills associated with a queue 125 may be the minimum set of skills that an agent 120 should posses to handle calls from the queue 125. The skills associated with a queue 125 may be set by a user or administrator.

To facilitate the assignment of agents 120 to queues, the environment 100 may further include a scheduler 170. The scheduler 170 may assign agents 120 to queues 125 based on the skills associated with the agents 120, the skills associated with the queues 125, and what is referred to herein as a “staffing” associated with each queue. The staffing associated with a queue 125 may be the minimum number of agents 120 that are needed to work on a queue 125 to maintain a particular service level. The service level may be defined by one or more metrics such as the maximum amount of time a customer 110 can be expected to wait to speak with an agent 120, for example. Other metrics may also be used.

The scheduler 170 may assign agents 120 to queues 125 for one or more intervals. An interval may be the smallest amount of time that an agent 120 can be scheduled for. Intervals used by the contact center 150 may be fifteen minutes, thirty minutes, forty-five minutes, or any appropriate time interval. The particular agents 120 assigned to a queue 125 for an interval is referred to herein as an “agent assignment.” The scheduler 170 may generate the staffing for a queue 125 for an interval based on a predicted workload for the queue 125 during the interval. The predicted workload may be based on historical workload data for the queue 125 and/or contact center 150 or may be provided by a user or administrator. Any method for predicting the workload of a queue 120 may be used.

The scheduler 170 may generate an agent assignment for a queue 125 for each interval based on the staffing generated for the queue 125 for the interval. For example, the call center 150 may use fifteen-minute intervals. The scheduler 170 may generate an agent assignment for the queue 125 for the 8:00 am interval based on the staffing for the interval, another agent assignment for the 8:15 am interval based on the staffing for the interval, and another assignment for the 8:30 am interval based on the staffing for the interval.

FIG. 2 illustrates a system architecture for incorporating a scheduler 170 into a business or entity such as a contact center 150. As shown the scheduler 170 includes various modules and components such as a skill group engine 210, a weight engine 220, and a schedule engine 230. More or fewer modules or components may be supported by scheduler 170. Each of the skill group engine 210, weight engine 220, and the schedule engine 230 may be implemented together or separately by one or more general purpose computing devices programmed with computer executable code that is stored in one or more non-transient memory devices. Furthermore, while shown as separate from the scheduler 170, in some implementations the scheduler 170 may be implemented as a component of the contact center 150.

The skill group engine 210 may divide or assign the agents 120 into skill groups 211. A skill group 211 may be a grouping of agents 120 based on the skills associated with each agent 120. In some implementations, each agent 120 associated with a skill group 211 may be associated with the same skills. The skills associated with a skill group 211 may be the skills associated with each of the agents 120 in the skill group 211. Each agent 120 may be assigned by the skill group engine 120 into only one skill group 211. Any method for assigning agents 120 into skill groups 211 may be used.

In some implementations, the skill group engine 210 may group agents 120 into skill groups 211 that have similar skills, rather than exactly the same skills. This type of grouping is referred to herein as a fuzzy skill group. For example, an agent 120 that is associated with the skills English and Spanish may be added to a skill group 211 associated with the skills English, Spanish, and Portuguese, even though the agent 120 does not speak Portuguese. Depending on the implementation, the skill group engine 210 may determine to “relax” skills that are not popular or that are not associated with many queues 125 in the contact center 150. Continuing the example above, the skill group engine 120 may have determined that the skill Portuguese is associated with very few queues 125, and/or the queues 125 that are associated with the skill Portuguese are not very busy or have low staffing 121.

After all of the agents 120 have been assigned to a skill group 211, the skill group engine 210 may further divide the skill groups 211 into what are referred to herein as networks 213. A network 213 may be a set of skill groups 211 where each skill group 211 in the network 213 has at least one skill in common with at least one other skill group 211 in the network 213. In addition, no skill group 211 in a first network 213 has any skill in common with any skill group 211 in a second network 213.

In some implementations, the skill group engine 210 may create a network 213 by selecting a skill group 211 for the network 213. The skill group engine 210 may determine the queue 125 that the agents 120 associated with the selected skill group 211 could work. Of the determined queues 125, the skill group engine 210 may determine the skill groups 211 whose agents 120 can work in the determined queues 125. These determined skill groups 211 may be added to the network 213. The skill group engine 210 may then continue adding skill groups 211 in this fashion until no more skill groups 211 can be added to the network 213.

The skill group engine 210 may then select a skill group 211 that has not yet been added to a network 213 and may create a network 213 using the selected skill group 211 as described above. As will be clearer based on the disclosure below, because none of the skill groups 211 in one network have any skills in common with the skill groups 211 in another network, the weight engine 220 may perform simulations and may calculate skill group weights 221 for the skill groups 211 in each network 213 in parallel.

The weight engine 220 may calculate a skill group weight 221 for each skill group 211 in a network 213 for each interval. As used herein, a skill group weight 221 for a skill group 211 may be a data structure that includes a weight for each skill associated with the skill group 211 for an interval. The weight for each skill may be based on how often an agent 120 from the skill group 211 worked on a task or communication that involves the skill during the associated interval. For example, if agent 120 in a skill group 211 spent 90% of their time in an interval working on the skill Spanish and 10% of their time in the interval working on the skill English, the skill group weight 221 for the skill group 211 for the interval would be 0.90 and 0.10.

In some implementations, the weight engine 220 may calculate the skill group weight 221 for a skill group 211, by running one or more simulations of the contact center 150. The simulation may be based on historical data for the contact center 150 and may simulate the customers 110, agents 120, and queues 125 associated with the contact center 150 for one or more intervals. Any method for simulating a contact center 150 may be used.

The weight engine 220 may determine from the simulations, how much time each agent 120 of the skill group 211 spent working using each of its skills during an interval. The determined times may be used by the weight engine 220 to determine a distribution of the agent's time across the skills during the interval. The distribution for each skill may be used as the weight for the skill for the interval. For example, If at 8 am on Monday the agent 120 spent 30% of his time on the skill English, 60% on the skill Spanish, and was idle 10% of the time (and the agent 120 is the only one in the skill group 211), the weight for the English skill during the interval Monday 8 am would be 0.333 (i.e., 30%/{30%+60%)) and the weight for the Spanish skill during the interval Monday 8 am would be 0.666 (i.e., 60%/{30%+60%)). Assuming the values are the same for every interval, the skill group weights 221 for three intervals for the skill group 211 of English and Spanish would be English (0.33, 0.33, 0.33) and Spanish (0.67, 0.67, 0.67).

Depending on the implementation, the skill group weight 221 for a skill group 211 during an interval may be determined by averaging the skill group weights 221 determined for each of the agents 120 in the skill group 211 over the interval. Note that in the event that a particular agent 120 does not do any work during a particular interval event though there was work to be done, in some implementations, the weights of the skill group weight 221 may be assigned by the weight engine 220 proportionally based on the workload of the queues that the other agents 120 in the skill group 211 worked. Depending on the implementation, the weights may be assigned such that the sum of the weights is always 1. Other methods for assigning the weights may be used.

The weight engine 220 may, after running each simulation of the contact center 150 for each interval, add up the number of agents 120 working on a particular skill weighted by the skill group weight 221 computed for their associated skill group 211. The computed number of agents 120 for each skill group 211 for each interval may provide a potential staffing curve for each skill that is referred to herein as “PS_SGW”. Note that in some implementations, agents 120 having only a single skill may not be considered when adding the number of agents 120 for each interval. As may be appreciated, if an agent 120 has only a single skill, then there may be no issue with determining how to divide the time of the agent 120.

Continuing the example above, in the simulation there may be two agents 120 having the skill English in a first interval, and three agents 120 having the skill English in the other two intervals. There may be one agent 120 having the skill Spanish in each of the three intervals. Assume there are four agents 120 in the skill group 211 of English and Spanish, and that the multi-skilled agents 120 were occupied 30/60/10 in intervals one and two (as discussed above), and occupied 60/20/20 in interval three (i.e., 0.75 in English, 0.25 in Spanish). Accordingly, the staffing curve PS_SGW for the three intervals would be English {1.32, 1.32, 3) and Spanish (2.68, 2.68, 1).

For real-time queues 125, the weight engine 220 may use reverse Erlang C, Erlang A, or a similar formula to compute the required staffing 231 for each skill and queue 125. Depending on the implementation, the required staffing 231 may be the number of agents 120 needed to work a queue 125 in order to meet a desired service level. The service level may be provided by one or more of the simulations ran by the weight engine 220.

The Erlang formula is a known mathematical equation for calculating the number of agents that is needed in a call center, given the number of calls and the desired service level to be achieved. The Erlang formula takes inputs like interaction volume, average handling time, and staffing and outputs a predicted service level. As may be appreciated, the weight engine 220 may reverse an Erlang formula to predict the staffing 231 required for the service level. For example, the weight engine 220 may use the service level provided by the simulation (along with the interaction volume and average handling time if available) and an Erlang formula to predict the staffing 231. The predicted staffing 231 for each interval may form a curve that is referred to herein as “PS_Erlang”. In order to account for single skilled agents 120, the weight engine 220 can remove these from PS_Erlang to generate a new Erlang staffing curve for just the multi-skilled agents 120. This curve is referred to herein as “PS MSE”.

The weight engine 220 may calculate the final skill group weights 221 for each skill group 211 by, for each skill group 211, adjusting each weight in the skill group weight 221 up by a percentage difference between the curves PS_SGW and PS_MSE for each interval for that skill. Continuing the example above, if PS_MSE was 20% higher than PS_SGW for the English skill in interval one, the final weight for English for the skill group 211 of English and Spanish in that interval would be 36% (i.e., 30%*1.2). Because this process is used to model the increasing effect of having a multi-skilled agent that can work on other queues 125 when one is idle, this process may be skipped for non-real-time queues 125.

As another example, the curve PS_Erlang may have the following weights for the skills English and Spanish of a skill group 221 for the intervals one, two, and three: PS_Erlang: English (3.5, 6, 5.5) and Spanish (3, 2, 3). The weight engine 220 may subtract the effect of the single skill agents 120 to get PS_MSE: English (1.5, 3, 2.5) and Spanish (2, 1, 2). From the example above, the value of PS_SGW for the intervals was PS_SGW: English {1.32, 1.32, 3) & Spanish (2.68, 2.68, 1), and the value of the skill group weight 221 for the intervals for the skill group 211 of English and Spanish was English (0.33, 0.33, 0.75) & Spanish (0.67, 0.67, 0.25).

The weight engine 220 may calculate the percent difference between PS_MSE and PS_SGW for each skill of the skill group weight 221 at each interval to get: English {1.13, 2.27, 0.83) and Spanish (0.75, 0.75, 2). Finally, the weight engine 220 may multiply the skill group weights 221 for the intervals by the differences to get the final skill group weights 221 for the intervals of: English (0.37, 0.75, 0.75) and Spanish (0.67, 0.67, 0.5).

In some implementations, when the weight engine 220 attempts to calculate a skill group weight 221 for a skill group 211 for a certain interval, during the simulation no agents 120 (or few agents 120) may have done any work with respect to some or all of the skills associated with the skill group 211 for that interval. Because no (or little) work was performed, it may be difficult for the weight engine 220 to determine the appropriate skill group weights 221 for the interval.

Depending on the implementation, the weight engine 220 may solve this problem in various ways. One solution is to find another interval having similar characteristics as the current interval. For example, the weight engine 220 may find an interval with a similar interaction volume or average handling time. The weight engine 220 may use the calculated skill group weight 221 for the skill group 211 for the similar interval for the current interval.

Another solution is to use a skill group weight 221 calculated for a similar skill group 211 for the same interval. For example, the weight engine 220 may select a skill group 211 with the most skills in common with the current skill group 211 and may determine the skill group weight 221 for the current skill group 221 based on the skill group weight 221 of the common skill group 211.

As another solution, the weight engine 220 may use the skill group weight 221 calculated for the current interval for a different simulation of the contact center 150 in the current set of simulations. Further, if so suitable skill group weight 221 is found in the current simulations for the current interval, the weight engine 220 may consider skill group weights 221 calculated for the same interval in past sets of simulations.

The weight engine 220 may attempt to find a suitable skill group weight 221 for the current interval using the methods described above. If no such skill group weight can be determined using any of the described methods, the weight engine 220 may use combinations of the above methods.

The skill group weights 221 may be calculated by the weight engine 220 periodically, and preferably before the skill group weights 221 are needed to place agents 120. As may be appreciated, simulating one or more queues 125 of a contact center 150 based on schedules and forecasts can be a very time consuming and resource intensive operation. Accordingly, the simulations may run periodically to generate the skill group weights 211, and the skill groups weights 211 may be later used when needed to place agents 120. This in an improvement over prior art systems that run simulations each time an agent 120 placement is needed, which is inefficient and results in delayed agent 120 placement.

The schedule engine 230 may use the calculated skill group weights 221 for each queue 125 for each interval to determine which queue 120 to place an agent 120 based on the skills associated with the agent 120. Depending on the implementation, the schedule engine 230 may receive a request to generate an agent assignment 233 for a set of queues 125 for one or more intervals. The agent assignment 233 may be an assignment of one or more agents 120 to the queues 125 of the contact center 150 for the one or more intervals.

As one example, the schedule engine 230 may receive a request for an agent assignment 233 of a plurality of agents 120 to a plurality of queues 125 for an interval. For each of some number of iterations, the schedule engine 230 may place the agents 120 into the queues 125 based on the skills required by each queue 125 and the skills associated with each agent 120 according to the required staffing of each queue 125 to generate an agent assignment 233.

After generating the assignment 233, the schedule engine 230 may calculate a score 235 for each of the queues 125 for the iteration. The score 235 for a queue 125 may be calculated based on the staffing 231 associated with the queue 125 and the skill group weights 221 associated with skill groups 211 of the agents 120 assigned to the queue 120. Depending on the implementation, the scores 235 may be calculated using a delta squared objective function. However, other functions may be used.

Generally, the schedule engine 230 may calculate a score 235 for a queue 125 for one or more intervals by, for each interval, taking the required staffing 231 for the interval minus the product of the number of agents 120 assigned to the queue 120 for the interval and the weight of the skill group weight 211 for the skill group 211 associated with the agents 120. The sum over each interval for the queue 125 may be the score for the queue 120.

For example, continuing the example from above. Assume a skill group weight 221 for the skill group 211 of English and Spanish for three intervals is English (0.37, 0.75, 0.75) and Spanish (0.67, 0.67. and 0.5). The schedule engine 230 may be calculating the score 235 for the placement of agents 120 from the skill group 211. There may be five agents 120 from the skill group 211 English and Spanish that may be placed in a queue 125 that has the required staffing 231 of one agent 120 with the skill English and two agents 120 with the skill Spanish for the first interval, five agents 120 with the skill English and one agent 120 with the skill Spanish for the second interval, and three agents 120 with the skill English and zero agents 120 with the skill Spanish for the third interval.

Already part of the agent assignment 233 for the three intervals may be agents 120 from the skill group 211 Spanish and agents from the skill group 211 English (i.e., single skill groups). In particular, there may be two agents 120 from the skill group 211 English and one agent 120 from the skill group 211 Spanish assign to work the first interval, there may be three agents 120 from the skill group 211 English and one agent 120 from the skill group 211 Spanish assign to work the second interval, and there may be three agents 120 from the skill group 211 English and one agent 120 from the skill group 211 Spanish assign to work the third interval.

The schedule engine 230 may calculate the score 235 for assigning the five agents 120 from the skill group 211 English and Spanish to the queue 125 using a delta squared objective function. In particular, the schedule engine 230 may calculate for each queue 125, and for each interval, the sum of the required agents 120 for each interval minus the number of agents 120 working times their skill group weight 221 for that skill. Thus, the score 235 for the queue 125 for the skill English would be:(1−(2+(5*0.37)))2+(5−(3+(5*0.75)))2+(3−(3+(5*0.75)))2=25

Similarly, the score 235 for the queue 125 for the skill Spanish would be:(2−(1+(5*0.67)))2+(1−(1+(5*0.67)))2+(0−(1+(5*0.5)))2=29

Accordingly, the total score 235 for the placement of the five agents 120 from the skill group 211 English and Spanish in the queue 125 for the three intervals would be 54.

After each of the iterations are completed, the schedule engine 230 may select the assignment 233 that received the overall best scores 235. Generally, the lower the score 235 the better the agent assignment 233 with respect to the associated queue 125. Accordingly, the schedule engine 230 may select the assignment that received the lowest total score across all of the queues 125.

FIG. 3 illustrates method 300 for dividing skill groups into a plurality of networks, and for calculating skill group weights for the skill groups in each network in parallel. The method 300 may be implemented by the scheduler 170.

At 310, information about a plurality of skill groups is received. The information may be received by the skill group engine 210 of the scheduler 170. Each skill group 211 may include one or more agents 120. In some implementations, the information may associate each skill group 211 with one or more skills. Each agent 120 may have some or all of the skills associated with its skill group 211.

At 315, the plurality of skill groups is divided into a first network and a second network. The plurality of skill groups 211 may be divided by the skill group engine 210. Each network 213 may include skill groups 211 that have no associated skills in common with any skill groups 211 in any other network 213. While only a first network 213 and a second network 213 are described, it is for illustrative purposes only; there is no limit to the number of networks 213 that may be supported.

At 320, for each skill group in the first network, a skill group weight is calculated. The skill group weights 211 may be calculated by the weight engine 220 for the same one or more intervals. As described above, the skill group weight 221 for a skill group 211 at an interval may be calculated by running simulations of the contact center 150 for the agents 120 in the skill group 211.

At 325, for each skill group in the second network, a skill group weight is calculated. The skill group weights 221 may be calculated by the weight engine 220 for the same one or more intervals. Because the first network 213 and the second network 213 have no skill groups 211 in common, the skill group weights 221 for the second network 213 may be calculated substantially in parallel with the skill group weights 221 for the first network 213.

FIG. 4 is an illustration of an example method 400 for generating and implementing an agent assignment 233 based on skill group weights 221. The method 400 may be implemented by the scheduler 170. At 401, a plurality of agent assignments is generated. The agent assignments 233 may be generated by the schedule engine 230 of the scheduler 170 for an interval by assigning agents 120 to queues 125 based on the skills associated with each agent 120 and the required staffing 231 needed to meet a desired service level for the interval. Any method for generating agent assignments 233 may be used.

At 403, a determination is made as to whether a simulation is required. The determination may be made by the weight engine 220. Depending on the implementation, the simulation of the contact center 150 and/or the queues 125 may be required when the interval has not yet been simulated by the weight engine 220, or a threshold amount of time has passed since a last simulation. If a simulation is required, the method 400 may continue at 405. Else, the method may continue at 409.

At 405, a simulation is ran. The contact center 150 may be simulated by the weight engine 220 for an interval. The contact center 150 may be simulated for the interval based on historical data about how busy the various agents 120 and queues 125 were handling communications for customers 110 of the contact center 150 for the same or similar intervals. Other information about the contact center 150 such as the IVO and AHT associated with the agents 120 may be used for the simulation. Depending on the implementation, the contact center 150 may be simulated multiple times for the interval. Any method for simulating a contact center 150 may be used.

At 407, skill group weights are calculated. The skill group weight 221 for each skill group 211 associated with an agent 120 may be calculated by the weight engine 220 using the results of the simulations. In some implementations, the weight engine 220 may calculate the skill group weight 221 for a skill group 211 by determining the amount of time that each agent 120 associated with the skill group 211 spent working on each associated skill. The determined amount of time for each skill may be used to determine the skill group weight 221. After calculating the skill group weights 221 for each skill group 211 for the interval, the skill group weights 221 may be stored for later use.

At 409, calculated skill group weights are retrieved. The calculated skill group weights 221 for the interval may be retrieved by the weight engine 220.

At 411, a score is calculated for the agent assignment. The score 235 for the agent assignment 233 may be calculated by the schedule engine 230. The score 235 may be calculated for the agent assignment 233 for the interval based on the skill group weights 221 associated with each skill group 211, the agents 120 assigned to each queue 120, and the required staffing 231 of the queues 125 for the interval. Depending on the implementation, the scores 235 may be calculated using a delta squared objective function.

At 413, a determination is made as to whether there are additional assignments to score. If there are additional assignments 233 to score, the method 400 may return to 403. Else, the method 400 may continue to 415.

At 415, the best assignment is selected based on the scores. The best agent assignment 233 may be selected by the schedule engine 230 of the scheduler 170. Depending on the implementation, the agent assignment 233 with the lowest (or highest) associated score 235 may be the best assignment 233. The selected assignment 233 may be implemented by the contact center 150 for the interval.

FIG. 5 shows an exemplary computing environment in which example implementations and aspects may be implemented. The computing system environment is only one example of a suitable computing environment and is not intended to suggest any limitation as to the scope of use or functionality. Numerous other general purpose or special purpose computing system environments or configurations may be used. Examples of well-known computing systems, environments, and/or configurations that may be suitable for use include, but are not limited to, personal computers, servers, handheld or laptop devices, multiprocessor systems, microprocessor-based systems, network personal computers (PCs), minicomputers, mainframe computers, embedded systems, distributed computing environments that include any of the above systems or devices, and the like. Computer-executable instructions, such as program modules, being executed by a computer may be used. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform particular tasks or implement particular abstract data types. Distributed computing environments may be used where tasks are performed by remote processing devices that are linked through a communications network or other data transmission medium. In a distributed computing environment, program modules and other data may be located in both local and remote computer storage media including memory storage devices.

With reference to FIG. 5, an exemplary system for implementing aspects described herein includes a computing device, such as computing device 500. In its most basic configuration, computing device 500 typically includes at least one processing unit 502 and memory 504. Depending on the exact configuration and type of computing device, memory 504 may be volatile (such as random access memory (RAM)), non-volatile (such as read-only memory (ROM), flash memory, etc.), or some combination of the two. This most basic configuration is illustrated in FIG. 5 by dashed line 506.

Computing device 500 may have additional features/functionality. For example, computing device 500 may include additional storage (removable and/or non-removable) including, but not limited to, magnetic or optical disks or tape. Such additional storage is illustrated in FIG. 5 by removable storage 508 and non-removable storage 510.

Computing device 500 typically includes a variety of tangible computer readable media. Computer readable media can be any available tangible media that can be accessed by device 500 and includes both volatile and non-volatile media, removable and non-removable media. Tangible, non-transient computer storage media include volatile and non-volatile, and removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data. Memory 504, removable storage 508, and non-removable storage 510 are all examples of computer storage media. Tangible computer storage media include, but are not limited to, RAM, ROM, electrically erasable program read-only memory (EEPROM), flash memory or other memory technology, CD-ROM, digital versatile disks (DVD) or other optical storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other medium which can be used to store the desired information and which can be accessed by computing device 500. Any such computer storage media may be part of computing device 500.

Computing device 500 may contain communications connection(s) 512 that allow the device to communicate with other devices. Computing device 500 may also have input device(s) 514 such as a keyboard, mouse, pen, voice input device, touch input device, etc. Output device(s) 516 such as a display, speakers, printer, etc. may also be included. All these devices are well known in the art and need not be discussed at length here.

Returning to FIG. 1, agent(s) 120 and customers 110 may communicate with each other and with other services over the network 130. For example, a customer calling on telephone handset may connect through the PSTN and terminate on a private branch exchange (PBX). A video call originating from a tablet may connect through the network 130 terminate on the media server. A smartphone may connect via the WAN and terminate on an interactive voice response (IVR)/intelligent virtual agent (IVA) components. IVR are self-service voice tools that automate the handling of incoming and outgoing calls. Advanced IVRs use speech recognition technology to enable customers to interact with them by speaking instead of pushing buttons on their phones. IVR applications may be used to collect data, schedule callbacks and transfer calls to live agents. IVA systems are more advanced and utilize artificial intelligence (AI), machine learning (ML), advanced speech technologies (e.g., natural language understanding (NLU)/natural language processing (NLP)/natural language generation (NLG)) to simulate live and unstructured cognitive conversations for voice, text and digital interactions. In yet another example, Social media, email, SMS/MMS, IM may communicate with their counterpart's application (not shown) within the contact center 150.

The contact center 150 itself be in a single location or may be cloud-based and distributed over a plurality of locations. The contact center 150 may include servers, databases, and other components. In particular, the contact center 150 may include, but is not limited to, a routing server, a SIP server, an outbound server, a reporting/dashboard server, automated call distribution (ACD), a computer telephony integration server (CTI), an email server, an IM server, a social server, a SMS server, and one or more databases for routing, historical information and campaigns.

The ACD is used by inbound, outbound and blended contact centers to manage the flow of interactions by routing and queuing them to the most appropriate agent. Within the CTI, software connects the ACD to a servicing application (e.g., customer service, CRM, sales, collections, etc.), and looks up or records information about the caller. CTI may display a customer's account information on the agent desktop when an interaction is delivered. Campaign management may be performed by an application to design, schedule, execute and manage outbound campaigns. Campaign management systems are also used to analyze campaign effectiveness.

For inbound SIP messages, the routing server may use statistical data from reporting/dashboard information and a routing database to the route SIP request message. A response may be sent to the media server directing it to route the interaction to a target agent 120. The routing database may include: customer relationship management (CRM) data; data pertaining to one or more social networks (including, but not limited to network graphs capturing social relationships within relevant social networks, or media updates made by members of relevant social networks); agent skills data; data extracted from third party data sources including cloud-based data sources such as CRM; or any other data that may be useful in making routing decisions.

The integration of real-time and non-real-time communication services may be performed by unified communications (UC)/presence sever. Real-time communication services include Internet Protocol (IP) telephony, call control, instant messaging (IM)/chat, presence information, real-time video and data sharing. Non-real-time applications include voicemail, email, SMS and fax services. The communications services are delivered over a variety of communications devices, including IP phones, personal computers (PCs), smartphones and tablets. Presence provides real-time status information about the availability of each person in the network, as well as their preferred method of communication (e.g., phone, email, chat and video).

The simulation discussed above generates events for each communication (call, email, chat, or the like) and agent staffing change. This effectively mimics the operation of the contact center: work items arriving, waiting in the queue, being handled by the agents, agents going on/off shift, . . . . To perform the simulation, estimations of the arrival numbers and average handling times of each work queue can be obtained, from an external provider for example. Also, agents scheduling information (e.g., shift intervals and breaks) can be obtained, from an external provider for example, The simulation can be organized in modules (e.g., software executing on computer hardware) that handle the events flowing through the contact center. for example, scheduler 170 can include a routing module that handles an item arriving and finding a suitable agent or being queued, and an agent tracking module that measures the agents capacity, availability and work queues they can handle.

For the simulation to be able to handle concurrent communications, it is necessary to track, for any given agent: 1) how many concurrent communications the agent can work on at the same time (max concurrent handling); 2) how many concurrent communications the agent is currently working on (current concurrent handling or percentage attention occupied as described in detail below); 3) if the agent is available to pick up concurrent and/or non-concurrent communications; 4) the queues that the agent can work.

During simulation, a concurrent communication will arrive at a queue based on the categorization of the communication. For example, if the communication relates to technical support, it will be assigned to a technical support queue. An agent to handle the communication is selected from a pool of available agents associated with the queue in the manner described above. The agent can be selected based on two criteria: 1) the agent that is currently engaged in the least number of concurrent interactions; and 2) if none is available, the agent that has been idle for the longest period of time. These criteria can be defined in an algorithm of a routing module of Scheduler 170. These criteria are just an example. The routing algorithm can include any appropriate logic for routing communications in a desired manner.

If no agent satisfies the logic of the routing algorithm, the item is queued, i.e., assigned to an appropriate waiting queue until an agent becomes available to work it (e.g., an existing agent associated with the queue finishes an item, or an agent associated with the queue starts a new shift). An agent is considered not available to take a new item if the agent is either handling a non-concurrent communication or is already at max capacity for concurrent items. The routing of non-concurrent items in the disclosed implementations can be the same as a conventional call center that does not manage concurrent interactions. As one example, in this implementation, the agent that is waiting for the longest time among the idle agents will be picked when a new non-concurrent item arrives (even if this agent is capable of handling concurrent work).

The simulation determines how much work volume the employees perform in the call center. Although scheduled, an agent is not always productive. For example, and agent may take bathroom breaks, encounter computer problems, have a meeting, or the like. This unproductive time, referred to as “shrinkage” herein, can be determined as a percentage of the total time worked by the agent or a group of agents and can be defined, for example, by a supervisor/administrator. Shrinkage can be defined with respect to any group of agents, such as agents assigned to a specific queue, agents on a team, individual agents or communication type. Shrinkage can also be defined for a specific queue (rather than agents assigned to a queue). In this case, an agent assigned to two queues can have two different shrinkage values in each queue.

During conventional simulation, shrinkage is modeled as an increase in the amount of time an agent spends on an item. This works well, but when calculating the Skill Group Weights (SGW), it is desirable to disregard shrinkage, so the shrinkage calculation is reversed. In a conventional environment, this reversal is straightforward to compute. However, when managing concurrent communications, an agent might handle two or more items of different work queues at the same time, each having a different shrinkage value. That is, if an agent worked concurrently on items from multiple queues, it must be determined how much of the time worked by the agent was productive non-shrinkage time.

To solve this, during simulation, the disclosed implementations calculate a weighted shrinkage based on the time worked in that interval. For example, with reference to FIG. 6, an example with two chat queues, C1 and C2, is illustrated. For this example, assume that C1 and C2 have shrinkage values of 10% and 20%, respectively. Also assume that Agent X has a max concurrent handling of 2 and is scheduled for 10 minutes. If a concurrent communication from C1 is assigned to Agent X at minute 2 and Agent X starts working on it and a concurrent communication from C2 is assigned to Agent X at minute 5, Agent X is now concurrently working on both communications. If the communication from C1 is finished at minute 6 (i.e., it took 4 minutes to be handled) and the communication from C2 is finished at minute 7 (i.e., took 2 minutes to be handled), Agent X works 4 minutes in C1, 2 minutes in C2, and the concurrent session took a total of 5 minutes (from minute 2 to minute 7). The weighted shrinkage on this work will be (4*0.1+2*0.2)/(4+2)=0.13. The actual time worked of this agent is then 5*(1−0.13)=4.35 minutes. Alternatively, as noted above, when routing communications, a concurrent communication can be assumed to take a percentage of the agent's attention, e.g., 20% for some queues and 50% for another. In this case, a communication can only be routed to an agent who is idle or is working on concurrent items and has a percent attention open which is greater than or equal to the percent attention required for the new concurrent communication. In this alternative, the max concurrent items used in the shrinkage calculator is computed as 1/(percent attention requires). For example, a queue that takes 20% of the agent's attention would have a max concurrent of 5.

After simulation, a disclosed implementation computes the initial SGW for a concurrent queue in an interval using the following algorithm:The SGW for concurrent queue C=(time on queue C/time on all concurrent items)*(elapsed concurrent time/non-idle time)

The “elapsed concurrent time” refers to the total physical time that elapsed while handling concurrent items; in contrast, the “time on queue A” and “time on all concurrent items” refer to the total time taken up by concurrent items. Therefore, if two items were being worked on at the same time, they would contribute twice the time to these values. Another example is illustrated in FIG. 7. In this example, Agent X worked on a communication from Voice queue V1 for 2 minutes, then a communication from Voice queue V2 for 3 minutes (both are non-concurrent communications). Then Agent X was assigned a chat communication (an immediate concurrent communication) item from chat queue C1 for 5 minutes and was Immediately assigned a second chat communication from queue C1 for 4 minutes. After working on the communications for queue C1 for 1 minute, Agent X was assigned a social media communication (a deferred concurrent communication) from queue C2 for 6 minutes. Agent X was then idle for the rest of the interval (3 minutes, in this example). Applying the the above algorithm to this example, yields initial SGWs of:(9/15)*(7/12)=0.35  C1(6/15)*(7/12)=0.233  C2

SGWs can be maintained for deferred concurrent queues but expanded to meet the reverse Erlang formula for immediate concurrent queues. In other words, for deferred concurrent queues, this is the final SGW value. For immediate concurrent queues, this value is expanded to meet the reverse Erlang formula. The difference between concurrent and non-concurrent in this example is that, when removing the single-skilled agents from the Erlang result to get the multi-skilled-Erlang component, that amount removed is increased by their max concurrent handling amount.

When scheduling agents, the SGW for a queue can be applied as the agent's effectiveness, multiplied by that agent's max concurrent handling value. To continue the example above, assume the scheduler schedules Agent X above at 8:00. Agent X's addition to the Staffing FTE for that interval would be 0.233*2=0.466 for C2. This contribution would be compared to the required staffing for that queue interval (or across several queue intervals in the case of a deferred queue) to compute a service goal score, in the manner described in U.S. application Ser. No. 16/668,525 for example. The contribution of Agent X to C1 would depend on the Erlang expansion, but for the purpose of simplicity, we can assume that there is no expansion needed and the contribution of Agent X to C1 is 0.35*2=0.7.

FIG. 8 illustrates the high-level logic flow 800 in accordance with a disclosed implementation. Before scheduling, a user configures Core Work Force Management (WFM) Services with agents, queues, scheduling rules, and any other scheduling information. Core WFM services 802 can be implemented by schedule engine 230 (FIG. 2). Historical call center data, such as call types, call volume and agent handling times (AHT) can be imported. A forecast algorithm of Core WFM services generates a prediction of likely future AHT and call volume. A staffing generator algorithm of Core WFM Services 802 creates staffing requirements for each queue based on that queue's forecasts. The staffing requirements, and any existing schedule are sent to a scheduler module such as scheduler 170 (FIG. 2).

To create a staffing schedule, Staff Differential Score Calculator 804 of the scheduler generates a score for the existing schedule (if any) based on staffing requirements and default/initial SGWs. The schedule score is sent to a search engine algorithm 806 which adjusts the schedule and recomputes the score using the SGWs. If the recomputed score is better, the adjusted schedule is used as the new current schedule. This staffing schedule calculation can be iterated plural times as needed to determine a potential schedule.

If simulation is required, based on a determination such as at 403 of FIG. 4 described above, simulator module 808, such as that described at 405 of FIG. 4 above, is executed based on the latest current schedule. If simulation is not required, the SGWs previously stored in SGW storage module 807 are used. During simulation, as each communication (e.g., call, chat, email, or the like) comes in during simulation, if there is not an agent available to be assigned to the communication, it is queued up until an agent becomes available or the communication is abandoned by the originator of the communication. If there's an agent available for that Queue, the agent takes the communication. If the communication is non-concurrent, that agent becomes unavailable for the duration of that communication. When the communication is completed, the agent becomes available to take a new communication. If the communication is a concurrent communication (e.g., a chat), if the agent has not yet reached their max concurrent items value, the agent can continue to take more concurrent communications, but not non-concurrent communications. Once the agent is concurrently handling as many items as the agent's max concurrent setting, that agent also becomes unavailable to take a new communication.

When any chat communication, or other non-concurrent communication, is completed by an agent who is at max concurrent, that agent becomes available to take new concurrent communications. When all concurrent communications are completed, the agent becomes available to take any communications (concurrent or non-concurrent).

During the simulation, statistics, such as percent service level, time spent on each queue, time spent on each type of communication by each agent, and the like are collected. After the simulation is finished, Skill Group Weight Calculator modules 810 computes new SGWs using the collected statistics (including the time spent working on each queue during the simulation). At this time, the logic can proceed to staffing differential score calculator 804, and search engine 806 again.

The disclosed implementations improve on the prior art by for example, computing SGWs for concurrently handled communications based on distribution of time worked which accounts for idle time and shrinkage and, in the case of immediate queues, expansion. Therefore, the implementations work for immediate (e.g., chat) and deferred (e.g., social media) concurrent communications.

The disclosed implementations can be implemented by various computing devices programmed with software and/or firmware to provide the disclosed functions and modules of executable code implemented by hardware. The software and/or firmware can be stored as executable code on one or more non-transient computer-readable media. The computing devices may be operatively linked via one or more electronic communication links. For example, such electronic communication links may be established, at least in part, via a network such as the Internet and/or other networks.

A given computing device may include one or more processors configured to execute computer program modules. The computer program modules may be configured to enable an expert or user associated with the given computing platform to interface with the system and/or external resources. By way of non-limiting example, the given computing platform may include one or more of a server, a desktop computer, a laptop computer, a handheld computer, a tablet computing platform, a Smartphone, a gaming console, and/or other computing platforms.

The various data and code can be stored in electronic storage devices which may comprise non-transitory storage media that electronically stores information. The electronic storage media of the electronic storage may include one or both of system storage that is provided integrally (i.e., substantially non-removable) with the computing devices and/or removable storage that is removably connectable to the computing devices via, for example, a port (e.g., a USB port, a firewire port, etc.) or a drive (e.g., a disk drive, etc.). The electronic storage may include one or more of optically readable storage media (e.g., optical disks, etc.), magnetically readable storage media (e.g., magnetic tape, magnetic hard drive, floppy drive, etc.), electrical charge-based storage media (e.g., EEPROM, RAM, etc.), solid-state storage media (e.g., flash drive, etc.), and/or other electronically readable storage media.

Processor(s) of the computing devices may be configured to provide information processing capabilities and may include one or more of a digital processor, an analog processor, a digital circuit designed to process information, an analog circuit designed to process information, a state machine, and/or other mechanisms for electronically processing information. As used herein, the term “module” may refer to any component or set of components that perform the functionality attributed to the module. This may include one or more physical processors during execution of processor readable instructions, the processor readable instructions, circuitry, hardware, storage media, or any other components.

Although the present technology has been described in detail for the purpose of illustration based on what is currently considered to be the most practical and preferred implementations, it is to be understood that such detail is solely for that purpose and that the technology is not limited to the disclosed implementations, but, on the contrary, is intended to cover modifications and equivalent arrangements that are within the spirit and scope of the appended claims. For example, it is to be understood that the present technology contemplates that, to the extent possible, one or more features of any implementation can be combined with one or more features of any other implementation.

While implementations and examples have been illustrated and described, it is to be understood that the invention is not limited to the precise construction and components disclosed herein. Various modifications, changes and variations may be made in the arrangement, operation and details of the method and apparatus disclosed herein without departing from the spirit and scope of the invention defined in the appended claims.

Claims

1. A method for scheduling agents in a call center to meet predefined service levels, wherein communications are associated with queues representing categories of communications, the queues including at least one concurrent queue of concurrent communications, wherein multiple concurrent communications can be handled concurrently by a single agent, the method comprising: executing a simulation to determine agent effectiveness of plural agents, the simulation including:computing a skill group weighting (SGW) for each agent for at least one concurrent queue and at least one interval based on: tc, the time spent by the agent on queue C communicationstall, the time spent by the agent on all concurrent communicationste, the elapsed concurrent time for the agenttn, the non-idle time of the agent; andscheduling the agents based on the SGW and max capacity of concurrent communications for each agent.

2. The method of claim 1, wherein the scheduling is based on the SGW multiplied by the max capacity of concurrent communications for each agent.

3. The method of claim 1, wherein shrinkage for each concurrent queue is calculated based on a percentage of total time worked in a queue for each agent.

4. The method of claim 1, wherein shrinkage for each concurrent queue is calculated based on a percentage of attention required in a queue for each agent.

5. The method of claim 1, wherein the scheduling includes calculating a staffing differential based on results of executing the simulation.

6. The method of claim 1 wherein computing a skill group weighting comprises applying the following equation: SGWC=(tc/tall)*(te/tn)

7. The method of claim 1 wherein at least one of the queues is an immediate concurrent queue.

8. The method of claim 1 wherein wherein at least one of the queues is a deferred concurrent queue.

9. The method of claim 1, wherein executing a simulation to determine an effectiveness of plural agents further comprises: receiving an estimation of the arrival numbers and average handling times of communications associated with each queue for at least one time interval;receiving a preliminary schedule of agents including shift intervals and breaks for each agent; andreceiving a max capacity of concurrent communications for each of the agents.

10. The method of claim 1, wherein shrinkage for each concurrent queue is calculated, the method further comprising; routing the communications to agents in accordance with a routing algorithm for each concurrent queue based on communications in the concurrent queue handled by each agent in the concurrent queue; andtracking capacity and availability for the agents for each queue.

11. A system for scheduling agents in a call center to meet predefined service levels, wherein communications are associated with queues representing categories of communications, the queues including at least one concurrent queue of concurrent communications, wherein multiple concurrent communications can be handled concurrently by a single agent, the system comprising: at least one computer hardware processor; andat least one memory device storing instructions which, when executed by the at least one processor, cause the at least one processor to carry out a method of: executing a simulation to determine an effectiveness of plural agents, the simulation including:computing a skill group weighting (SGW) for each agent for at least one concurrent queue and at least one interval based on: tc, the time spent by the agent on queue C communicationstall, the time spent by the agent on all concurrent communicationste, the elapsed concurrent time for the agenttn, the non-idle time of the agent; andscheduling the agents based on the SGW and max capacity of concurrent communications for each agent.

12. The system of claim 11, wherein the scheduling is based on the SGW multiplied by the max capacity of concurrent communications for each agent.

13. The system of claim 11, wherein shrinkage for each concurrent queue is calculated based on a percentage of total time worked in a queue for each agent.

14. The system of claim 11, wherein shrinkage for each concurrent queue is calculated based on a percentage of attention required in a queue for each agent.

15. The system of claim 11, wherein the scheduling includes calculating a staffing differential based on results of executing the simulation.

16. The system of claim 11 wherein computing a skill group weighting comprises applying the following equation: SGWC=(tc/tall)*(te/tn)

17. The system of claim 11 wherein at least one of the queues is an immediate queue.

18. The system of claim 11 wherein wherein at least one of the queues is a deferred queue.

19. The system of claim 11, wherein executing a simulation to determine an effectiveness of plural agents further comprises: receiving an estimation of the arrival numbers and average handling times of communications associated with each queue for at least one time interval;receiving a preliminary schedule of agents including shift intervals and breaks for each agent; andreceiving a max capacity of concurrent communications for each of the agents.

20. The system of claim 11, wherein shrinkage for each concurrent queue is calculated, the method further comprising; routing the communications to agents in accordance with a routing algorithm based on a weighted shrinkage for each concurrent queue based on communications in the concurrent queue handled by each agent in the concurrent queue; andtracking capacity and availability for the agents for each queue.

21. Non-transitory computer-readable media having instructions stored thereon which, when executed by a computer processor, cause the computer processor to carry out the method comprising: executing a simulation to determine an effectiveness of plural agents, the simulation including:computing a skill group weighting (SGW) for each agent for at least one concurrent queue and at least one interval based on: tc, the time spent by the agent on queue C communicationstall, the time spent by the agent on all concurrent communicationste, the elapsed concurrent time for the agenttn, the non-idle time of the agent; andscheduling the agents based on the SGW and max capacity of concurrent communications for each agent.