Predictive internet automatic work distributor (Pre-IAWD) and proactive internet automatic work distributor (Pro-IAWD)

• US 6,718,330 B1
• Filed: 12/16/1999
• Issued: 04/06/2004
• Est. Priority Date: 12/16/1999
• Status: Expired due to Term

• Current Assignee: NCR Voyix Corporation
• Original Assignee: NCR Corporation
• Inventors: Zenner, Hal B.
• Primary Examiner(s): LE, UYEN T
• Application Number:
  US09/465,003
• Time in Patent Office: 1,573 Days
• Field of Search: 707/10, 707/1, 707/2, 707/6, 707/104.1, 709/206, 709/217, 709/218, 709/219, 709/103, 709/104, 709/105
• US Class Current: 707/694
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising
• CPC Class Codes: Y10S 707/99945 Object-oriented database structure processing

Abstract

A Predictive IAWD (Pre-IAWD) initially assigns inbound work to an agent in a group of agents based on a dynamically generated prediction of work completion time. The Pre-IAWD initially assigns the work to the agent having the highest likelihood of completing the work first in the group of agents, based on a prediction generated by a Pre-IAWD algorithm based on dynamically measured factors, such as a current work load of each agent, a priority of the work, and an experience level and skill set of each agent. A Proactive IAWD (Pro-IAWD) reassigns work initially assigned to agents by the Pre-IAWD based on a set of predetermined rules, to maintain an equitable work load across the agent group.

Claims

1. A computer implemented method of distributing on-line client requests to human and/or robotic Internet service agents, comprising the following steps:

receiving a request from a client;

determining an average time each agent in a group of agents takes to service a request;

selecting an initial agent most likely to finish servicing the request first from a group of agents;

assigning the request to the initial agent; and

reassigning an assigned request from the initial agent to a second agent selected from the group of agents;

wherein said selecting step comprises;

determining a priority and a topic of the request by parsing the request, and

queuing the request for service by the initial agent ahead of previous requests having lower priorities; and

wherein said selecting step comprises, for each agent in a group of agents,

predicting an availability time when the agent will become available to service the request.

2. The method of claim 1, wherein said selecting step comprises comparing the availability and average times of each agent against the availability and average times of each of the other agents in the group of agents to determine which agent is most likely to finish servicing the request first.

3. The method of claim 1, wherein said selecting step comprises, for each agent in the group of agents, determining an elapsed time the agent has spent servicing a respective current request;

and determining an average time the agent takes to service the respective current request.

4. The method of claim 2, wherein said step of determining the average time the agent takes to service the request includes the steps of:

determining an experience level of the agent;

determining a skill set of the agent; and

comparing the skill set to the topic of the request.

5. The method of claim 1, wherein said selecting step comprises escalating the priority of the request after the request is assigned to an agent if the agent does not begin servicing the request within a predetermined time after the request is assigned to the agent.

6. The method of claim 1, wherein said selecting step includes selecting the initial agent based on at least one of a number of available agents, a skill set of each agent, an experience level of each agent, a number of requests queued for each agent, a running average time it takes for each agent having a specific skill set to service predetermined types of requests including information requests and repair requests.

7. The method of claim 1, further comprising at least one of triggering the reassigning step at periodic intervals in response to a periodic trigger event, and triggering the reassigning step when a work load of the initial agent exceeds a current average work load for each agent computed across the group of agents by a predetermined amount.

8. The method of claim 1, wherein the reassigning step comprises selecting one or more rules from a set of rules and reassigning the assigned request based on the one or more selected rules.

9. The method of claim 8, wherein the reassigning step comprises setting a precedence level among two or more selected rules such that one of the selected rules takes precedence over the remaining selected rules.

10. The method of claim 8, wherein said reassigning step comprises reassigning the request based on at least one of a total number of requests currently queued for and not being serviced by each agent, a priority level of each request currently queued for each agent, an experience level of each agent, and determining which agent is most likely to finish servicing the assigned request first.

11. The method of claim 1, wherein said reassigning step comprises for each agent:

determining a difference between a current work load of the agent and a current average work load of each agent computed across the group of agents;

comparing the determined difference against a predetermined threshold amount; and

reassigning the assigned request based on the results of each of said comparing steps.

12. The method of claim 11, wherein the current work load of each agent is based on one of the priorities of respective requests assigned to the agent, a number of requests assigned to the agent, and an average time it takes the agent to completely service a specific type of request including an information request and a repair request.

13. A computer implemented method of distributing on-line client requests to human and/or robotic Internet service agents, comprising the following steps:

determining an average time each agent in a group of agents takes to service a request;

receiving a request from a client and assigning the request to an initial agent in a group of agents; and

reassigning the request from the initial agent to a subsequent agent selected from the group of agents in response to a trigger event;

wherein said selecting step comprises;

determining a priority and a topic of the request by parsing the request, and

queuing the request for service by the initial agent ahead of previous requests having lower priorities; and

wherein said selecting step comprises, for each agent in a group of agents,

predicting an availability time when the agent will become available to service the request.

14. The method of claim 13, comprising one of periodically triggering said reassigning step and algorithmically triggering said reassigning step, said algorithmically triggering step including generating said trigger event when a work load of the first agent exceeds a current average work load for each agent computed across the group of agents by a predetermined amount.

15. The method of claim 13, wherein the reassigning step comprises:

selecting one or more rules from a set of rules, and, if two or more rules are selected, setting a precedence level among two or more of the selected rules such that one of the selected rules takes precedence over the remaining selected rules, and

selecting the subsequent agent from the group of agents using the selected rules, said one or more rules being base on at least one of a total number of requests currently queued for and not being serviced by each agent, a priority level of each request currently queued for each agent, an experience level of each agent, and determining which agent is most likely to finish servicing the request first.

16. A computer architecture, comprising:

receiving means for receiving a request from a client;

determining means for determining an average time each agent in a group of agents takes to service a request;

selecting means for selecting an initial agent most likely to finish servicing the request first from a group of agents;

assigning means for assigning the request to the initial agent; and

reassigning means for reassigning an assigned request from a first agent to a second agent selected from the group of agents;

wherein said selecting means comprises;

another determining means for determining a priority and a topic of the request by parsing the request, and

a queuing means for queuing the request for service by the initial agent ahead of previous requests having lower priorities; and

wherein said selecting means comprises, for each agent in a group of agents,

a predicting means for predicting an availability time when the agent will become available to service the request.

17. An article including a computer readable medium having stored thereon a plurality of sequences of instructions, said plurality of sequences of instructions including sequences of instructions which, when executed by a processor, cause said processor to perform the steps of:

receive a request from a client;

determine an average time each agent in a group of agents takes to service a request;

select an initial agent most likely to finish servicing the request first from a group of agents;

assign the request to the initial agent; and

reassign an assigned request from a first agent to a second agent selected from the group of agents;

wherein said selecting step comprises;

determining a priority and a topic of the request by parsing the request, and

queuing the request for service by the initial agent ahead of previous requests having lower priorities; and

wherein said selecting step comprises, for each agent in a group of agents,

predicting an availability time when the agent will become available to service the request.

18. A computer system, comprising:

a processor; and

a memory coupled to said processor, the memory having stored therein sequences of instructions which, when executed by said processor, cause said processor to;

receive a request from a client;

determine an average time each agent in a group of agents takes to service a request;

select an initial agent most likely to finish servicing the request first from a group of agents;

assign the request to the initial agent; and

reassign an assigned request from a first agent to a second agent selected from the group of agents;

wherein said selecting an initial agent comprises;

determining a priority and a topic of the request by parsing the request, and

queuing the request for service by the initial agent ahead of previous requests having lower priorities; and

wherein said selecting an initial agent comprises, for each agent in a group of agents,

predicting an availability time when the agent will become available to service the request.