Method of integrating artificial intelligence systems to dynamically adjust wrap-up time

Abstract

The present disclosure provides, among other things, a method of managing a wrap-up time in a contact center, the method including: receiving, by an agent of the contact center, a communication having a variable associated with the communication; receiving an input from a source external to the contact center; determining that the variable is related to the input; based on the relation of the variable to the input, determining an updated wrap-up time; storing the updated wrap-up time and the input in a database including timing variables; enabling a machine learning process to analyze the database; providing the updated wrap-up time to the agent as an amount of time rendered on a display to the agent; and updating a data model used to automatically determine wrap-up times based on the analysis of the machine learning process.

Description

315 325 includes a historical wrap-up database 386, a wrap-up timing decision database 380, a wrap-up/communication inputs 388, a wrap-up timing engine 382, and wrap-up information 384.

The learning module 374 may utilize machine learning and have access to training data and feedback 378 to initially train behaviors of the learning module 374. Training data and feedback 378 contains training data and feedback data that can be used for initial training of the learning module 374 and this training data may be different than, and is not to be confused with, training data for training agents, which is also discussed herein. The learning module 374 may also be configured to learn from other data, such as further contact center events or message exchanges based on feedback, which may be provided in an automated fashion (e.g., via a recursive learning neural network) and/or a human-provided fashion (e.g., by one or more human agents, such as agents 131A-131N). The learning module 374 may additionally utilize training data and feedback 378. For example, the learning module 374 may have access to one or more data model(s) 376 and the data model(s) 376 may be built and updated by the learning module 374 based on the training data and feedback 378. The data model(s) 376 may be provided in any number of formats or forms. Non-limiting examples of data model(s) 376 include Decision Trees, Support Vector Machines (SVMs), Nearest Neighbor, and/or Bayesian classifiers.

The learning module 374 may also be configured to access information from a wrap-up timing decision database 380 for purposes of building a historical wrap-up database 386. The wrap-up timing decision database 380 stores data related to wrap-up timings, including but not limited to historical wrap-up timings, wrap-up processing history, wrap-up notes, agent(s) (e.g., one or more of agents 131A-131N) notes associated with wrap-up timings, wrap-up decisions, etc. Wrap-up information within the historical wrap-up database 386 may constantly be updated, revised, edited, or deleted by the learning module 374 as the wrap-up engine 325 processes additional wrap-up timing decisions.

In some embodiments, the wrap-up engine 325 may include a wrap-up timing engine 382 that has access to the historical wrap-up database 386 and selects appropriate wrap-up processing decisions (e.g., presented in wrap-up information 384) based on input from the historical wrap-up database 386 and based on wrap-up/communication inputs 388 received from the call handler 321 and/or external information 372. The wrap-up engine 325 may receive wrap-up/communication inputs 388 in the form of external information 372 (e.g., social media network(s) 140, sentiment data 141, news feed(s) 142, and/or business rule(s) 143, etc.), real-time wrap-up data from the call handler 321, and/or other communication or contact center 120 information from the call handler 321. For example, the wrap-up/communication inputs 388 may include information about any wrap-up timing(s) recently set for various communication sessions, and variations or violations of wrap-up timings, and any other information from the call handler 321. The wrap-up timing engine 382 may make decisions regarding what wrap-up information 384 to provide to update wrap-up times based on any of the criteria described herein, including but not limited to information provided by the learning module, information about current status of wrap-up times (e.g., violations of current wrap-up times), feedback about wrap-up times (e.g., from agents, supervisors, agent training information, etc.), performance metrics related to wrap-up times or any other aspects of the contact center 120, customer feedback, etc. To enhance capabilities of the wrap-up timing engine 382, the wrap-up timing engine 382 may constantly be provided with training data and feedback 378 from conversations between agents 131A-131N and customers that have occurred over a particular timeframe, communication channel, location, and/or other parameters. Therefore, it may be possible to train a wrap-up timing engine 382 to have a particular output or multiple outputs. In various embodiments, the output of an AI application (e.g., learning module 374) is an updated wrap-up time that is sent, with appropriate information conveying how to apply the updated wrap-up time to specified communications, via the wrap-up timing engine 382 and from wrap-up information 384, to the wrap-up API 370.

Using the wrap-up/communication inputs 388 and the historical wrap-up database 386, the wrap-up engine 325 may be configured to provide wrap-up information 384 (e.g., one or more wrap-up times) to the wrap-up API 370 so that the wrap-up API 370 can publish one or more events, based on the wrap-up information 384, to the call handler 321 to update one or more wrap-up times for one or more communication sessions. An event of an updated wrap-up time may be provided as an output of the wrap-up timing engine 382 (e.g., processed through the wrap-up API 370) to the call handler 321. The API may be a readily visible API that is exposed to an external system (e.g., a system external to the contact center 120 such as the wrap-up engine 325) and the API injects an event (e.g., a Wrap Up event) into the contact center system (e.g., the call handler 321) to modify a wrap-up value on an ad-hoc basis based on the determinations (e.g., the wrap-up information 384) of the external system.

In some embodiments, the event resulting from the invocation of the API is published onto a Kafka bus of the contact center 120 and a Ready Service has subscribed for interest in wrap-up events. The Ready Service is, for example, a microservice that determines whether an agent is available for new work by consuming events (e.g., including disconnect events from engagements such as web chats, social media, and voice/video calls) from a Kafka bus. An event may be published on the Kafka bus by invoking an Agent State API, which permits authorized clients to influence the states of the contact center 120 agents. Thus, a wrap-up event can be consumed by the Ready Service, which then uses the time identified in the wrap-up event to set the wrap-up time for the applicable communication that is in-progress or about to end. In various embodiments, wrap-up processing logic may be included within the Ready Service, a wrap-up event may be included in an event catalogue, and a wrap-up API may be introduced into the Agent State API grouping. Further, external applications, for example, an AI application that is monitoring variables (such as Real Time Sentiment and Customer Net Promoter Score (also known as “NPS,” which is a measure of the satisfaction of a customer or how likely a customer is to recommend you to a friend or colleague)) can be authorized to invoke the wrap-up API. Therefore, wrap-up time as disclosed herein becomes an event-based data item, which can be determined in real time and exposed to authorized applications to influence.

In various embodiments, there can be little or no manual configuration of the wrap-up times and no static value for the wrap-up times because the wrap-up times may be set on an ad hoc basis. For example, an AI or machine learning application may be enabled to integrate with the systems and methods of the contact center 120 in order to advantageously determine and implement updated and customizable wrap-up times. Such embodiments are advantageous by automating and quickly adjusting (with little or no manual configuration) wrap-up times in the contact center 120 so that resources of the contact center 120 are saved.

In some embodiments, the wrap-up engine 325 may interact with the contact center 120 as follows. The call handler 321 may serve a plurality of agent terminals 130A-130N and there can be a plurality of communications occurring between the call handler 321 and the user endpoints 101A-101N. Event notifications are created by scripts in the contact center 120 and/or the wrap-up engine 325. The scripts in the wrap-up engine 325 may be run by the wrap-up timing engine 382 using information from one or more of the wrap-up/communication inputs 388, the historical wrap-up database 386, and/or the learning module 374. The scripts can create events that are posted to the bus in the call handler 321, and the events may map communication reference information in each event to a particular communication or group of communications (e.g., by a identifying a particular agent terminal, user endpoint, skill group 124, contact center queue 123, etc.), so that any actions performed based on the event can be appropriately implemented to the desired communication or set of communications. The set of communications may be all communications involving any one or more agents, agent terminals 130A-130N, users, user endpoints 101A-101N, skill group 124, contact center queue 123, etc. The scripts may create events that are directed to a common resource (e.g., a database, a contact center queue 123, a skills group 124, etc.) that multiple agents use, or events that are directed to a common property (e.g., a geographical location, a specific business name, a user identification, a key word and/or key phrase, etc.), and such events may require different variations in mapping. Thus, events concerning updated wrap-up times are entirely customizable, and have the ability to be communication-specific, be directed to a common resource and/or property, etc.

Referring now to FIG. 4, a first communication method will be described in accordance with at least some embodiments of the present disclosure. The method begins when a communication is received at step 402. The communication may be from a user via a communication device (e.g., one or more of communication endpoints 101A-101N) and received in a contact center 120. The communication may be any type of communication described herein (e.g., text-based communication, chat communication, SMS communication, webcast communication, email communication, voice communication, video communication, etc.). In various embodiments, the communication can be received at a wrap-up engine 125 at a same time as, or after, it is received at the contact center 120.

It is determined if a wrap-up time should be adjusted at step 406. In various embodiments, an external system such as a wrap-up engine 125 can make this determination based on historical data (e.g., related to the contact center 120, one or more users, external information, etc.), communication inputs, current external or internal information, etc. If the wrap-up time is not being adjusted, then the method proceeds to step 410 and the wrap-up time is set. The wrap-up time may be set, in part, by storing the wrap-up time in a memory accessible by a component of the contact center 120, such as the wrap-up engine 125. Alternatively, the wrap-up time may be set by injecting a Wrap Up event to a bus of the contact center 120, where the bus publishes events related to the communication for other components of the contact center to digest and act upon. In other words, in some instances, the wrap-up time may be set by not submitting any event (e.g., to a call handler 121 of the contact center 120) so that a previously determined static wrap-up time remains in place. In other embodiments, the wrap-up time may be set by submitting an event to be published that contains a set wrap-up time for the communication that is not updated by the wrap-up engine 125.

If the wrap-up time is adjusted at step 406, then the method proceeds to determine the updated wrap-up time in step 414. For example, an external system, such as a wrap-up engine 125, may determine what information should be used to update the wrap-up time. Such a decision may be based on information related to the contact center 120, one or more users, external and/or internal information, communication inputs, etc. The updated wrap-up time may be determined by a system that is external or internal to the contact center 120, and may be the wrap-up engine 125 as described herein in combination with an API that creates the wrap-up event and publishes the wrap-up event on a bus of the contact center 120 as described herein.

As one example of an updated wrap-up time, an input received by the wrap-up engine 125 could be related to a rule in the contact center that limits the wrap-up time to a maximum wrap-up time. This could be a business rule sent inside of the contact center. So, if the wrap-up engine 125 judged a call to be highly emotional (e.g., based on sentiment data), then the wrap-up engine 125 could determine that an agent needs an updated-wrap up time that is an extended amount of time. The wrap-up engine 125 could then invoke an API to generate the event with the extended wrap-up time and the event would hit the component within the contact center 120 that manages the wrap-up time values. However, the business rules would be considered when the updated wrap-up time is implemented so that the wrap-up time would be limited to the maximum wrap-up time.

Once the updated wrap-up time is determined, the wrap-up time is set at step 410. The wrap-up time may be set, in part, by storing the wrap-up time in a memory accessible by a component of the contact center 120, such as the wrap-up engine. Alternatively, the wrap-up time may be set by injecting a Wrap Up event to a bus of the contact center 120, where the bus publishes events related to the communication for other components of the contact center to digest and act upon. In other words, the wrap-up time may be set by submitting an event (e.g., to a call handler 121 of the contact center 120) to be published that contains an updated wrap-up time for the communication.

The method may further determine whether the communication has ended at step 418. If the communication has not ended, the method may proceed back to step 406 to further determine whether the wrap-up time should be adjusted 406. To determine whether the wrap-up time should be adjusted at step 406, the considerations provided herein may be considered, together with any updated information that has occurred since the previous time that it was determined whether to adjust the wrap-up time at step 406. For example, there may be relevant external information, or new information about or within the communication and the contact center 120, to consider.

If the communication has ended at step 418, then the wrap-up time may be applied at step 422. For example, the wrap-up time may be applied by displaying the wrap-up time on a screen of the agent to convey the wrap-up time to the agent. The wrap-up time may be displayed during the communication, updated one or more times during the communication, and after the communication ends. The wrap-up time may count down after the communication ends to convey the passing of the wrap-up time to the agent.

Referring now to FIG. 5, a second communication method will be described in accordance with at least some embodiments of the present disclosure. The method begins when input is received from a learning module at step 504. The input may be received by a system that is internal to the contact center 120 or external to the contact center 120, such as the wrap-up engine 325. The input is related to a communication of the contact center 120. The input may be information that is based on various sources of information, including but not limited to training data and feedback 378, data model(s) 376, and information from wrap-up timing decision database 380. One or more of the various sources of information may have enabled the learning module to access and process the various sources of information so that the learning module can provide the input to the system.

At step 508, a first update to the wrap-up time is determined. This first update may be determined by a wrap-up timing engine 382 as described herein. The first update is to a wrap-up time of the communication from step 504 and the first update may be saved to memory and/or communicated to an API. The first updated wrap-up time may be published; for example, the API may process the information provided by a system (e.g., an external system such as the wrap-up engine 325) and communicate a first updated wrap-up time event to a bus of the contact center 120 to be published on the bus.

At step 512, input is received from one or more external sources. The input may be received by the same system as step 508, which may be the wrap-up engine 325. The input is related to the same communication as in step 504. The input may be information that is based on various external sources, including but not limited to social media network(s) 140, sentiment data 141, news feed(s) 142, and/or business rule(s) 143, etc.

Based upon the input from the external source in step 512, a second update to the wrap-up time for the communication is determined at step 516. The second update may be determined to be a same wrap-up time as previously determined or set for the communication, or a different amount of time for the wrap-up time for the communication. The second update may be saved to memory and/or communicated to the API. The second updated wrap-up time may be published; for example, the API may process the information provided by a system (e.g., an external system such as the wrap-up engine 325) and communicate the second updated wrap-up time event to a bus of the contact center 120 to be published on the bus. Thus, the second update to the warp-up time may be published at step 520.

Referring now to FIG. 6, a third communication method will be described in accordance with at least some embodiments of the present disclosure. The method begins when an event is received that a communication is in progress at step 604. The communication being in progress means that an agent has been connected to a customer of the contact center 120, and the agent and the customer are engaged in a communications session. The event may be published to a bus of the contact center 120.

An updated wrap-up time is received at step 608. The updated wrap-up time may be received by a wrap-up timing engine 380 from a learning module 374 and the updated wrap-up time is related to the communication session from step 604. The updated wrap up time is published at step 612. For example, the updated wrap-up time may be published to the bus of the contact center 120.

An end of communication event is received at step 616. This may be received as a published event on the bus, for example. The end of communication event may be digested by an application that also digests updated wrap-up time events and applies wrap-up times to communications. Thus, in step 620, the updated wrap-up time may be set (e.g., applied) to the communication. When the updated wrap-up time is set, it can be displaying as a time value to the agent handling the communication session. The time value can count down to provide the agent with a visual indication (e.g., a countdown) of the amount of time associated with the updated wrap-up time.

At step 624, a violation of the wrap-up time is received. The violation may occur when an agent pauses or exceeds the updated wrap-up time, which may be published as an event on the bus. The violation may be received by a system that is internal or external to the contact center, such as the wrap-up engine 325. For example, the violation may be provided to wrap-up/communication inputs 388 and then to wrap-up timing engine 382 where it is passed on to wrap-up timing decision database 380 and on to learning module 374. The violation may be processed by the learning module 374 in combination with other properties of the communication session, other communications sessions, and other information that is internal or external to the contact center 120, to be used in future determinations of wrap-up times in the contact center 120.

The violation of the updated wrap-up time is reported in step 628. For example, the violation may be reported to internal or external systems of the contact center 120 that track metrics related to the contact center 120. Also, the violation may be reported to a supervisor of the agent handling the communication session. The report, or properties of the report, may be provided to the wrap-up engine 325, such as an input to the wrap-up/communication inputs 388. At step 632, the updated wrap-up time ends.

The present disclosure, in various aspects, embodiments, and/or configurations, includes components, methods, processes, systems, and/or apparatus substantially as depicted and described herein, including various aspects, embodiments, configurations embodiments, subcombinations, and/or subsets thereof. Those of skill in the art will understand how to make and use the disclosed aspects, embodiments, and/or configurations after understanding the present disclosure. The present disclosure, in various aspects, embodiments, and/or configurations, includes providing devices and processes in the absence of items not depicted and/or described herein or in various aspects, embodiments, and/or configurations hereof, including in the absence of such items as may have been used in previous devices or processes, e.g., for improving performance, achieving ease and\or reducing cost of implementation.

The foregoing discussion has been presented for purposes of illustration and description. The foregoing is not intended to limit the disclosure to the form or forms disclosed herein. In the foregoing Detailed Description for example, various features of the disclosure are grouped together in one or more aspects, embodiments, and/or configurations for the purpose of streamlining the disclosure. The features of the aspects, embodiments, and/or configurations of the disclosure may be combined in alternate aspects, embodiments, and/or configurations other than those discussed above. This method of disclosure is not to be interpreted as reflecting an intention that the claims require more features than are expressly recited in each claim. Rather, as the following claims reflect, inventive aspects lie in less than all features of a single foregoing disclosed aspect, embodiment, and/or configuration. Thus, the following claims are hereby incorporated into this Detailed Description, with each claim standing on its own as a separate preferred embodiment of the disclosure.

Moreover, though the description has included description of one or more aspects, embodiments, and/or configurations and certain variations and modifications, other variations, combinations, and modifications are within the scope of the disclosure, e.g., as may be within the skill and knowledge of those in the art, after understanding the present disclosure. It is intended to obtain rights which include alternative aspects, embodiments, and/or configurations to the extent permitted, including alternate, interchangeable and/or equivalent structures, functions, ranges or steps to those claimed, whether or not such alternate, interchangeable and/or equivalent structures, functions, ranges or steps are disclosed herein, and without intending to publicly dedicate any patentable subject matter.

Claims

1. A method of managing a wrap-up time in a contact center, the method comprising:

receiving, by an agent of the contact center, a communication having a variable associated with the communication;

receiving an input from a source external to the contact center;

determining that the variable is related to the input;

based on the relation of the variable to the input, determining an updated wrap-up time;

storing the updated wrap-up time and the input in a database comprising timing variables;

enabling a machine learning process to analyze the database;

providing the updated wrap-up time to the agent as an amount of time rendered on a display to the agent; and

updating a data model used to automatically determine wrap-up times based on the analysis of the machine learning process.

2. The method of claim 1, further comprising:

receiving input from the agent during the wrap-up time, wherein the receiving the input comprises at least one of storing a result and automatically reporting a result; and

displaying an end of the updated wrap-up time to the agent.

3. The method of claim 1, wherein the input from the external source is received during the communication.

4. The method of claim 1, wherein the machine learning process retrieves data related to the variable from the timing variables database, and wherein the updated wrap-up time is based on the data.

5. The method of claim 1, wherein the machine learning process performs an analysis of a sentiment contained in the communication.

6. The method of claim 1, wherein the data model is used to automatically classify communications, and further comprising providing a second updated wrap-up time to an agent associated with a second communication based on the automatic classification and the input, wherein the second updated wrap-up time is a second amount of time rendered on the display to the agent.

7. The method of claim 1, wherein the variable changes during the communication and wherein the machine learning process performs the analysis over a timeframe that includes the changes in the variable.

8. The method of claim 1, wherein the updated wrap-up time is determined based on a rule of the contact center considered in combination with the analysis by the machine learning process.

9. The method of claim 1, further comprising:

receiving a secondary input from a second source external to the contact center, wherein the second source is one of a news feed, a social media feed, and a business rule, and wherein the updated wrap-up time is determined by the machine learning process and based on both of the input and the secondary input.

10. The method of claim 1, wherein the machine learning process analyzes the database to obtain the input before a start of the communication.

11. A communication system, comprising:

a processor; and

computer memory storing data thereon that enables the processor to:

receive, by an agent of the contact center, a communication having a variable associated with the communication;

receive an input from a source external to the contact center;

determine that the variable is related to the input;

based on the relation of the variable to the input, determine an updated wrap-up time;

store the updated wrap-up time and the input in a database comprising timing variables;

enable a machine learning process to analyze the database;

provide the updated wrap-up time to the agent as an amount of time rendered on a display to the agent; and

update a data model used to automatically determine wrap-up times based on the analysis of the machine learning process.

12. The communication system of claim 11, wherein the processor is further enabled to receive the input from the agent during the wrap-up time, wherein the receiving the input further comprises at least one of storing a result and automatically reporting a result.

13. The communication system of claim 11, wherein the input from the external source is received during the communication.

14. The communication system of claim 11, wherein the machine learning process retrieves data related to the variable from the timing variables database, and wherein the updated wrap-up time is based on the data.

15. The communication system of claim 11, wherein the machine learning process is further enabled to determine the updated wrap-up time based on a sentiment contained in the communication.

16. The communication system of claim 11, wherein the data model is used to automatically classify communications, and the processor is further enabled to provide a second updated wrap-up time to an agent associated with a second communication based on the automatic classification and the input.

17. The communication system of claim 11, wherein the processor is further enabled to invoke a reporting function when the updated wrap-up time is provided to the agent or when the data model is updated.

18. The communication system of claim 11, wherein the updated wrap-up time is determined based on a rule of the contact center considered in combination with the analysis by the machine learning process.

19. The communication system of claim 11, wherein the processor is further enabled to receive a secondary input from a second source external to the contact center, wherein the second source is one of a news feed, a social media feed, and a business rule, and wherein the updated wrap-up time is determined by the machine learning process and based on both of the input and the secondary input.

20. A contact center, comprising:

a server comprising a processor and a wrap-up timing engine that is executable by the processor and that enables the processor to:

receive, by an agent of the contact center, a communication having a variable associated with the communication;

receive an input from a source external to the contact center;

determine that the variable is related to the input;

based on the relation of the variable to the input, determine an updated wrap-up time;

store the updated wrap-up time and the input in a database comprising timing variables;

enable a machine learning process to analyze the database;

provide the updated wrap-up time to the agent as an amount of time rendered on a display to the agent; and

update a data model used to automatically determine wrap-up times based on the analysis of the machine learning process.

21. A method of managing an updated wrap-up time in a contact center, the method comprising:

receiving, by an agent of the contact center, a communication having a variable associated with the communication;

receiving an input;

determining that the variable is related to the input;

based on the determined relation of the variable to the input, determining an updated wrap-up time of the communication;

storing the updated wrap-up time in a database comprising timing variables; and

providing a status of the updated wrap-up time to the agent on a display to the agent.

22. The method of claim 21, wherein the variable is received during the communication.

23. The method of claim 21, further comprising:

receiving input from the agent during the updated wrap-up time, wherein the receiving the input comprises at least one of storing a result and automatically reporting a result.

24. The method of claim 21, wherein the status comprises displaying a countdown of the updated wrap-up time.

25. The method of claim 21, wherein the variable is related to a sentiment.

26. The method of claim 25, wherein the sentiment is identified during the communication.

27. The method of claim 21, further comprising:

analyzing the variable to classify the communication based on a rule regarding a sentiment.

28. The method of claim 21, wherein the variable changes during the communication, and wherein the updated wrap-up time is based on the variable changes.

29. The method of claim 21, wherein the variable changes during the communication, and wherein a second updated wrap-up time is determined based on the variable changes.

30. The method of claim 29, further comprising:

providing a status of the second updated wrap-up time to the agent on the display.

31. The method of claim 30, wherein the status comprises a countdown of the second updated wrap-up time on the display, and wherein the countdown starts when the communication ends.

32. A method of managing a wrap-up time in a contact center, the method comprising:

receiving, by an agent of the contact center, a communication having a variable associated with the communication;

receiving an input from a source external to the contact center, wherein the input is related to the variable;

based on the determined relation of the variable to the input, determining an updated wrap-up time of the communication;

storing the updated wrap-up time in a database comprising timing variables; and

providing the updated wrap-up time to the agent as an amount of time rendered on a display to the agent.

33. The method of claim 32, wherein the input is received during the communication.

34. The method of claim 32, wherein the determining the updated wrap-up time is after determining that the input is related to the variable.

35. The method of claim 32, further comprising:

retrieving data related to the variable from the database, wherein the updated wrap-up time is based on the data.

36. The method of claim 32, further comprising:

analyzing the database to obtain the input.

37. A system, comprising:

a processor; and

computer memory storing data thereon that enables the processor to:

receive, by an agent of a contact center, a communication having a variable associated with the communication;

receive an input;

determine that the variable is related to the input;

based on the determined relation of the variable to the input, determine an updated wrap-up time of the communication;

store the updated wrap-up time in a database comprising timing variables; and

provide a status of the updated wrap-up time to the agent on a display to the agent.

38. The system of claim 37, wherein the processor is further enabled to receive input from the agent during the updated wrap-up time.

39. The system of claim 38, wherein the receiving the input further comprises at least one of storing a result and automatically reporting a result.

40. The system of claim 37, wherein the processor is further enabled to retrieve data related to the variable from the database, and wherein the updated wrap-up time is based on the data.