System and method for managing a dynamic call flow during automated call processing

Abstract

A system and method for managing a dynamic call flow during automated call processing is provided. Calls from a plurality of callers into a call center are accommodated through monitored call sessions. A call flow for each call session is managed. Caller characteristics of the caller and physical conditions in relation to the call are evaluated. Those caller characteristics and physical conditions, which are respectively unique among the callers and the calls, are identified. The call flow of each call session is dynamically adjusted by modifying parameters for one or more of the caller characteristics and physical conditions. The caller characteristics, physical conditions, and parameters are stored in a database.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a functional block diagram showing an automated call center operational environment.

FIG. 2 is a functional block diagram showing an automated call center for use in the automated call center operational environment of FIG. 1.

FIG. 3 is a process flow diagram showing, by way of example, a typical call sequence, as transacted by the automated call center of FIG. 2.

FIG. 4 is a screen diagram showing, by way of example, a view of an agent console for a live call session.

FIG. 5 is a process flow diagram showing a method for managing a dynamic call flow during automated call processing, in accordance with one embodiment.

FIG. 6 is a block diagram showing, by way of example, caller attributes evaluated during dynamic call flow management.

FIG. 7 is a block diagram showing, by way of example, call flow parameters customized during dynamic call flow management.

FIG. 8 is a block diagram showing a system for managing a dynamic call flow during automated call processing, in accordance with one embodiment.

DETAILED DESCRIPTION

Automated Call Center Operational Environment

Automated call centers are a form of voice-response system, which rely upon programmed control to guide callers through a machine-generated dialogue for receiving support and problem resolution assistance. FIG. 1 is a functional block diagram showing an automated call center operational environment 10. By way of example, a multiplicity of callers call into an automated call center 11, generally through telephonic means, which can include conventional telephone handsets 13a-c through Plain Old Telephone Service (POTS) 12, portable handsets 15a-c through cellular and satellite telephone service 14, VoIP clients 17a-b, and Internet telephony clients 18a-b, as well as other forms and combinations of telephony, message, and voice-based communications. For example, a telephone caller may provide responses to the automated call center 11 through a combination of voice, dual-tone multi-frequency (DTMF), or text messaging. Callers can also “call” or interface into the automated call center 11 using conventional network clients 19 through an internetwork 17, including the Internet. Other types of automated call center access are possible.

Except as otherwise stated, as used herein, the terms “caller,” “user,” and “customer” are used interchangeably to refer to a caller to the automated call center 11. Similarly, the terms “agent,” “guide,” and “operator” are used interchangeably to refer to an agent that provides service provisioning to the caller for the automated call center 11. Additionally, although the automated call center 11 is shown as a single point within the automated call center operation environment 10, the automated call center 11 could include two or more logically interconnected but physically separate, including geographically removed, call center operations, which collectively provide a unified automated call center. Other automated call center arrangements and configurations are possible.

Automated Call Center

Generally, an automated call center provides a single dial-in source for support and problem resolution assistance for customers seeking direct help from manufacturers and service vendors. However, automated call centers 11 can also be applied to many other diverse areas of commerce, such as mail order sales and roadside assistance dispatch. FIG. 2 is a functional block diagram showing an automated call center 30 for use in the automated call center operational environment 10 of FIG. 1. The automated call center 30 includes one or more servers 31, 34 and agent consoles 32a-c operatively interconnected over an internal network 33. The agent consoles 32a-c could also be interconnected to the message server 31 over an external network infrastructure 33, such as the Internet or a non-public enterprise data network. The network infrastructure 33 can be either wired or wireless and, in one embodiment, is implemented based on the Transmission Control Protocol/Internet Protocol (TCP/IP) network communications specification, although other types or combinations of networking implementations are possible. Similarly, other network topologies and arrangements are possible.

A message server 31 provides the primary message-based communications infrastructure for the call center operation, such as described in commonly-assigned U.S. Patent Publication No. 2003/0177009, published Sep. 18, 2003 to Odinak et al., and U.S. Patent Publication No. 2005/0177368, published Aug. 11, 2005 to Odinak, the disclosures of which are incorporated by reference. During regular operation, the message server 31 executes multiple threads to process multiple simultaneous calls, which are handled by agents executing agent applications on agent consoles 32a-c.

Customer calls are received through a telephony interface 35, which is operatively coupled to the message server 31 to provide access to a telephone voice and data network 36. In one embodiment, the telephony interface connects to the telephone network 36 over a T-1 carrier line, which can provide individual channels of voice or data traffic. Other types of telephone network connections are possible.

The automated call center 30 also includes a recognition server 34, which interfaces directly to the message server 31 as a top-level or root tier of a speech recognition hierarchy. The message server 31 sends streamed audio data for each user call to the recognition server 34, which then performs distributed speech recognition. The message server 31 assigns grammar generation and speech recognition to the recognition server 34. Upon startup, the telephony gateway 35 opens a T-1 carrier device channel for each available T-1 time slot. The telephony gateway 35 initiates a new connection to the message server 31, one connection per T-1 device channel, and the message server 31, in turn, initiates a corresponding new connection to the recognition server 34.

The separate telephony gateway-to-message server and message server-to-main recognizer connections form one concurrent session apiece. When a customer call is answered or connected, the telephony gateway 35 sends a call message to the message server 31. The message server 31 then sends a new call message to the recognition server 34.

The message server 31 also assigns each new call to one of the agent consoles 32a-c through dynamic load balancing, such as further described in commonly-assigned U.S. Provisional Patent application, entitled “System and Method for Balancing Agent Console Load During Automated Call Processing,” Ser. No. 60/838,074, filed Aug. 15, 2006, the disclosure of which is incorporated by reference. Each agent console 32a-c provides the primary means for direct customer interaction by providing service provisioning and related assistance to callers, such as further described below with reference to FIG. 3. The main purpose of each agent console 32a-c is to execute one or more agent applications, which display both user and agent messages and provide menus of actions that can be executed in response to agent commands, including script execution. One or more agent applications execute on each agent console and one or more agent consoles can execute in parallel. Alternatively, multiple instances of agent applications can run on a server machine and can be accessed by agents at agent consoles operating as remote terminals.

Each component, including the message server 31, recognition server 34, and agent consoles 32a-c, is implemented as a computer program, procedure or module, or state transition written as source code or flow specification in a conventional programming language, such as the C++ programming language or modeling tool, and presented for execution by a computer system as object or byte code. Alternatively, the components could be directly implemented in hardware, either as integrated circuitry or burned into read-only memory components. The various implementations of the source code and object and byte codes can be held on a computer-readable storage medium.

Exemplary Typical Call Sequence

Callers to the automated call center interface indirectly with an agent, who multitasks through one or more individual caller sessions via an agent console 32a-c. The callers, however, only perceive an automated calling interface heard, for instance, through their telephone or portable handset. Calls are processed through a sequence of phases. FIG. 3 is a process flow diagram showing, by way of example, a typical call sequence 40, as transacted by the automated call center 30 of FIG. 2. Although the form of call processing required may vary by subject matter area and other factors, the same overall sequence of caller-to-agent interchanges will loosely apply.

Initially, upon calling into the automated call center 30, each user receives an initial greeting and optionally informational message 41 providing a synopsis of caller options. When provided, the caller options enable the user to navigate through to a specific topic area for assistance or support. However, the user could simply proceed directly into a customer support scenario 42 with either an agent, which is a live person; with a guide, which is an automated prompt that is under the supervision of a live person; or through automated voice response to enable information collection, problem trouble-shooting, and other actions. When present, the guide can listen to and revise caller replies in the background, generally unbeknownst to the caller, such as described in commonly-assigned U.S. Patent Application Publication No. 2003/0177009, entitled “System and Method for Providing a Message-Based Communications Infrastructure for Automated Call Center Operation,” published Sep. 18, 2003, pending, the disclosure of which is incorporated by reference. Other guide functions and caller dispositions are possible.

The scenario 42 can be delayed by agent unavailability, caller volume capacity limits, and other factors that can delay providing a response to the caller. Throughout the session, the agent can adjust and customize the flow of the call processing by evaluating caller attributes, such as described below beginning with reference to FIG. 5. In addition, the order in which information is gathered from the user to assist with problem resolution, such as identifying the caller's needs or troubleshooting a problem, can be dynamically evaluated and controlled, such as further described in commonly-assigned U.S. Provisional Patent application, entitled “System and Method for Processing Out-Of-Order Caller Responses During Automated Call Processing,” Ser. No. 60/838,102, filed Aug. 15, 2006, the disclosure of which is incorporated by reference.

As required, service provisioning 43 is provided to the user, either directly during the course of the call or indirectly through a service request dispatch. The type of service provided can include sending information or products, entering a subscription or enrollment, or other forms of commerce or assistance that can be remotely provided over the telephone. Finally, the call ends in a wrap-up 44, which provides closure to the call and a departing salutation. Other forms and variations of customer call sequences are feasible. Importantly, however, from the perspective of the caller, the experience may appear to be an interaction with an intelligent machine and the caller would thus be aware that the agent is automated, not human. Accordingly, the caller would have a more relaxed expectation of agent responsiveness since a machine, and not an actual person, is on the line.

The automated call center 30 can perform in-progress call processing 46, which operates on a stream of speech utterances copied from on-going calls. As well, following call termination, the automated call center 30 can perform post-call processing 45, which identifies individual speech utterances in each call that can be analyzed or stored into a database 34. The identified speech utterances can be presented to an agent for manipulation, such as ranking or reordering. In addition, the post-call processing 45 can include performing speech recognition on the speech utterances, identifying speaker characteristics, and marking certain speech utterances for later use. In-progress 46 and post-call 45 processing can use the same types of operations, but in-progress call processing 46 allows an agent to perform real time call analysis.

Exemplary Agent Console

Call sessions are displayed on each agent console through a graphical user interface (GUI). FIG. 4 is a screen diagram showing, by way of example, a view of an agent console 50 for a live call session. Multiple call sessions can be simultaneously displayed and managed through the GUI controls.

The particulars concerning the live call session are obtained by the message server 31 and provided to the agent. The session particulars can include, for instance, caller profile 51, address 52, contact information 53, reason for call 54, and credit card data 55. Other types of session particulars are possible.

Operationally, the agent can review a recording of the caller's spoken responses through intuitive playback controls 57 and update the display by operating an “Update” control 56. Additionally, the agent can transfer the caller to a live agent by operating a “Transfer to Live Agent” control 58 and can log out by operating a “Logout” control 59. Other GUI controls are possible.

Dynamic Call Flow Management

Caller sessions can be dynamically customized by an agent through an agent console 32a-c at any point during a call. FIG. 5 is a process flow diagram showing a method 70 for managing a dynamic call flow during automated call processing, in accordance with one embodiment. Although a caller interface would ordinarily be modified or adjusted by the agent that is assigned to a call session, the caller interface could also be dynamically modified by another agent console or a server based on those caller attributes and conditions that are known. For example, the message server 31 may independently increase the sound volume for a caller through the telephony interface 35 if the message server 31 detects a significant degradation in line quality for that caller. Other types of dynamic caller interface changes are possible.

Generally, the caller and agent engage in interaction throughout the call session (operation 71), such as described above with reference to FIG. 3. Throughout the call session, the agent, with the assistance of the agent console 32a-c, will evaluate the attributes of the caller (operation 72) and can customize the flow of the call (operation 73), as further described below respectively with reference to FIGS. 6 and 7. Caller interaction, attribute evaluation, and call flow customization can continue to be performed repeatedly as needed throughout the session.

Caller Attributes

Caller attributes depend upon both the caller and the physical characteristics of the call itself. FIG. 6 is a block diagram showing, by way of example, caller attributes 81 evaluated 80 during dynamic call flow management. Caller-specific attributes relate to perceived personal idiosyncrasies of the caller. Physical characteristics exist independently from the caller and are artifacts of both the physical medium and timing of the call.

Caller attributes 81 that relate to the specific caller include emotion 82, accent 83, gender 84, and age 85. Unless independently confirmed, for instance, through a questionnaire or database lookup, these caller attributes are perceived based on the subjective impressions of the agent or through automated analysis.

Physical caller-independent attributes 81 include origin of the call 86 and the time of the call 87. The caller origin 66 can be determined, for example, through area code or zip code lookup, and the time of the call 87 is available through automated timestamp. Other caller attributes 88 are possible.

Call Flow Parameters

Based on the caller attributes evaluated, the caller interface can be customized dynamically to adjust the flow of the call to better meet the needs of the caller. FIG. 7 is a block diagram showing, by way of example, call flow parameters 91 customized 90 during dynamic call flow management. Call flow can be adjusted either directly by the agent console 32a-c of the agent responsible for the call, or indirectly through another agent console or server.

The parameters adjustable to customize the call flow 91 include language 92, flow or detail level of the call 93, word choice 94, and speed 95. The content of the call can be further customized by tailoring those products that are recommended to the caller 96. Additionally, based primarily upon caller-specific attributes 81, the recognition grammar used 97 by the recognition server 34 (shown on FIG. 2) can be changed to better suit the caller's needs. Other call flow parameters 98 are possible.

Agent Console

Call flow is primarily controlled by the agent console 32a-c for the agent assigned to handling the call session. FIG. 8 is a block diagram showing a system 100 for managing a dynamic call flow during automated call processing, in accordance with one embodiment. The system 100 operates in accordance with a sequence of process steps, as described above with reference to FIG. 5.

Throughout the call session, the agent receives a stream of messages 109 and generates a stream of responses 100 through his or her agent console 101. The console 101 includes a script engine 102 that the agent can use to execute stored scripts 106 maintained in a storage device 105, which present “canned” interaction dialogues to the caller.

The console 101 also includes an evaluator 103 and controller 104. Throughout each session, the evaluator 103 parses and identifies caller attributes 107, which are stored into the storage device 105 and evaluated. Finally, the controller 104 suggests or automatically selects parameters 108 to dynamically adjust the call flow based on the evaluated attributes 107. The parameters 108 are also stored in the storage device 105. Other console functionality is possible.

While the invention has been particularly shown and described as referenced to the embodiments thereof, those skilled in the art will understand that the foregoing and other changes in form and detail may be made therein without departing from the spirit and scope of the invention.

Claims

1. A system for managing a dynamic call flow during automated call processing, comprising: a display to monitor a call session of a caller;an evaluator to evaluate caller attributes in relation to the caller during the call session and to identify those caller attributes, which are unique to at least one of the caller and the call session; anda controller to dynamically adjust a flow of the call session by modifying parameters corresponding to the unique caller attributes.

2. A system according to claim 1, further comprising one or more of: an agent adjustment module to adjust the call session flow through prompting via an agent; anda dynamic adjustment module to dynamically adjust the call session flow through at least one of a server and agent console.

3. A system according to claim 2, further comprising: a window to display the call session on the agent console; andan update module to maintain the agent console to provide user information to an agent.

4. A system according to claim 1, wherein the evaluator parses the caller attributes from the call session, further comprising: a selection module to select those parameters most closely associated with the parsed caller attributes.

5. A system according to claim 1, wherein the parameters comprise at least one of language, flow, word choice, speed, products, and recognition grammar.

6. A method for managing a dynamic call flow during automated call processing, comprising: monitoring a call session of a caller;evaluating caller attributes in relation to the caller during the call session;identifying those caller attributes, which are unique to at least one of the caller and the call session; anddynamically adjusting a flow of the call session by modifying parameters corresponding to the unique caller attributes.

7. A method according to claim 6, further comprising one or more of: adjusting the call session flow through prompting via an agent; anddynamically adjusting the call session flow through at least one of a server and agent console.

8. A method according to claim 7, further comprising: displaying the call session on the agent console; andmaintaining the agent console to provide user information to an agent.

9. A method according to claim 6, further comprising: parsing the caller attributes from the call session; andselecting those parameters most closely associated with the parsed caller attributes.

10. A method according to claim 6, wherein the parameters comprise at least one of language, flow, word choice, speed, products, and recognition grammar.

11. A system for managing a dynamic call flow during automated call processing, comprising: a recognition server to accommodate calls from a plurality of callers into a call center through monitored call sessions;an agent console to manage a call flow for each call session, comprising: an evaluator to evaluate caller characteristics of the caller and physical conditions in relation to the call and to identify those caller characteristics and physical conditions, which are respectively unique among the callers and the calls; anda controller to dynamically adjust the call flow of each call session by modifying parameters corresponding to one or more of the caller characteristics and physical conditions; anda database to store the caller characteristics, physical conditions, and parameters in a database.

12. A system according to claim 11, further comprising: a decision module to determine the caller characteristics based on at least one of subjective impressions of an agent and an automated analysis.

13. A system according to claim 11, wherein the caller characteristics comprise at least one of emotion, accent, gender, and age.

14. A system according to claim 11, wherein the physical conditions comprise at least one of a call origin and time.

15. A system according to claim 11, further comprising: a call processor to process the call session at least one of during the call session and after the call session to identify speech utterances.

16. A method for managing a dynamic call flow during automated call processing, comprising: accommodating calls from a plurality of callers into a call center through monitored call sessions;managing a call flow for each call session, comprising: evaluating caller characteristics of the caller and physical conditions in relation to the call;identifying those caller characteristics and physical conditions, which are respectively unique among the callers and the calls; anddynamically adjusting the call flow of each call session by modifying parameters corresponding to one or more of the caller characteristics and physical conditions; andstoring the caller characteristics, physical conditions, and parameters in a database.

17. A method according to claim 16, further comprising: determining the caller characteristics based on at least one of subjective impressions of an agent and an automated analysis.

18. A method according to claim 16, wherein the caller characteristics comprise at least one of emotion, accent, gender, and age.

19. A method according to claim 16, wherein the physical conditions comprise at least one of a call origin and time.

20. A system according to claim 16, further comprising: processing the call session at least one of during the call session and after the call session to identify speech utterances.