DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Response to Amendment
The amendment filed on July 10th, 2026, has been entered. Claims 1, 12, and 19 have been amended. Claims 25 and 26 have been added. Claims 1-13, 15-16, 19-20, and 23-26 are pending and have been examined.
Response to Arguments
Applicant’s arguments with respect to claim(s) 1-13, 15-16, 19-20, and 23-24 have been considered but are moot because the new ground of rejection does not rely on any reference applied in the prior rejection of record for any teaching or matter specifically challenged in the argument.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claim(s) 1 and 5-8 is/are rejected under 35 U.S.C. 103 as being unpatentable over McCord, Alan (US Patent No. 11,741,484 B2 hereinafter McCord), in view of Kumar et al. (US Pat. Pub. No. 2018/0146093 A1 hereinafter Kumar), Neuer et al. (US Patent No. 10,237,405 B1 hereinafter Neuer), and Sivasubramanian et al. (US Pat. Pub. No. 2021/0158805 A1 hereinafter Sivasubramanian).
Regarding claim 1, McCord discloses a computer-implemented method for outputting feedback to a selected device in a call center (McCord, Col. 11, lines 61-66: “FIG. 3 is a block diagram illustrating an exemplary system architecture for a contact center utilizing a self-learning interaction optimizer (SLIO) 300 comprising a reinforcement learning server 310 and an optimization server 320 (both shown below in FIG. 4), according to one aspect.”), the method comprising: accessing behavioral and lexical features determined from audio data associated with a conversation between an agent of the call center and a caller (McCord, Col. 17, lines 5-23: "According to various aspects, when an agent (i.e., a contact center agent) interacts with a customer (typically a caller into a contact center, or a customer who has been called by the contact center), a series of measurements of semantic or other distance between the language of the agent and the language of the customer is made"; "“Distance” is used here as a term analogous to “conversational similarity”, and distance measurements may be made based on semantic distance, stylistic distance, emotional distance, acoustic difference (for example, differences in tonal range and prosody/rhythm), or other similar distance measures appropriate for measuring distance between spoken and/or written text fragments."). However, McCord fails to expressly recite collecting, from a device associated with the agent, information currently being displayed on the device; accessing, from a customer relationship management (CRM) system, customer relationship management (CRM) data that includes input from the agent, the information displayed on the device associated with the agent, management flow data associated with the conversation, information about the caller, a current stage of a current workflow being used for the conversation, and information about the agent; applying the behavioral and lexical features and the CRM data to one or more models that classify aspects of the conversation, wherein the one or more models are structured as a plurality of processing elements comprising a directed acyclic graph executed as a parallel algorithm in which each processing element of the plurality of processing elements performs an operation as a parallel process when a set of inputs becomes available to the one or more models, and wherein the one or more models were trained to provide information on a current state of the conversation using a plurality of labeled conversations; receiving, from the one or more models, guidance data and scoring data determined based at least partially on the behavioral and lexical features and the CRM data, wherein the guidance data includes guidance for the agent in the conversation with the caller, and the scoring data includes a rating of the conversation; outputting to the CRM system, a notification comprising the guidance data and scoring data in a format associated with the CRM system, wherein the guidance data comprises a suggested new stage of the current workflow for the conversation; determining that additional audio data associated with the conversation is available since the notification was generated; in response to the determination, performing behavioral and lexical analysis on the additional audio data; generating feedback data related to the conversation based, at least in part, on the transmission of the notification and the behavioral and lexical analysis on the additional audio data, and output the feedback data to the device associated with the agent in the conversation during the conversation, wherein the guidance data comprises the suggested new stage of the current workflow for the conversation.
Kumar teaches collecting, from a device associated with the agent, information currently being displayed on the device; and accessing the information displayed on the device associated with the agent (Kumar, [0027]: "The recording of the communications session may be performed in addition, or as an alternative, to routing the communications session to human agents for handling. The recording may be performed in one or both of two distinct ways: (1) the customer's interaction with the human agent (e.g., an audio recording of the speech-based conversation between the customer and the agent during a voice call), or the interactive response system, or both may be recorded, and/or (2) the computer screens presented to the human agent by the agent's computer during the agent's interaction with the customer may be recorded or captured (often referred to as “screen captures”).").
McCord and Kumar are analogous arts because they both belong to the same field of endeavor
of customer service systems. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the customer interaction and experience system of McCord to incorporate the teachings of Kumar to capture information from the screen of an agent. Recording what is being displayed on an agent’s screen can help monitor the agent’s effectiveness (Kumar, [0028]). This ensures that any issues with an agent’s effectiveness can be properly addressed. However, McCord, in view of Kumar, fails to expressly recite accessing, from a customer relationship management (CRM) system, customer relationship management (CRM) data that includes input from the agent, management flow data associated with the conversation, information about the caller, a current stage of a current workflow being used for the conversation, and information about the agent; applying the behavioral and lexical features and the CRM data to one or more models that classify aspects of the conversation, wherein the one or more models are structured as a plurality of processing elements comprising a directed acyclic graph executed as a parallel algorithm in which each processing element of the plurality of processing elements performs an operation as a parallel process when a set of inputs becomes available to the one or more models, and wherein the one or more models were trained to provide information on a current state of the conversation using a plurality of labeled conversations; receiving, from the one or more models, guidance data and scoring data determined based at least partially on the behavioral and lexical features and the CRM data, wherein the guidance data includes guidance for the agent in the conversation with the caller, and the scoring data includes a rating of the conversation; outputting to the CRM system, a notification comprising the guidance data and scoring data in a format associated with the CRM system, wherein the guidance data comprises a suggested new stage of the current workflow for the conversation; determining that additional audio data associated with the conversation is available since the notification was generated; in response to the determination, performing behavioral and lexical analysis on the additional audio data; generating feedback data related to the conversation based, at least in part, on the transmission of the notification and the behavioral and lexical analysis on the additional audio data, and output the feedback data to the device associated with the agent in the conversation during the conversation, wherein the guidance data comprises the suggested new stage of the current workflow for the conversation.
Neuer teaches accessing, from a customer relationship management (CRM) system, customer relationship management (CRM) data that includes (Neuer, Col. 18, line 60 - Col. 19, line 5: "The CARM may also provide the meta-data related to the call to the file store 190 via signaling 235. For example, for each call stored in the file store 190, the CARM may provide meta-data relating to timing information of checkpoints that occurred during the call. The timing information, as will be seen, may be defined in various formats. One format may be an offset time from the beginning of each call for each checkpoint. For example, the meta-data may have an offset time #1 for the first checkpoint, an offset time #2 for the second checkpoint, etc. The file store 190 may store this information in association with the call recording, so that the call recording and the meta-data can be retrieved at a later time for reviewing the checkpoint widget."): input from the agent (Neuer, Col. 19, lines 2-5: "The file store 190 may store this information in association with the call recording, so that the call recording and the meta-data can be retrieved at a later time for reviewing the checkpoint widget."; Here, the call recording includes speech input from the agent.), management flow data associated with the conversation (Neuer, Fig. 14A; Col. 33, lines 42-48: "In this embodiment, the checkpoints 1410, 1420, 1430, 1440, 1450 are arranged as a series of checkboxes along with text in a list format. There is an implied time-line reflecting the passage of time in a downward manner (e.g., the first checkpoint associated with the first line is expected to occur before the second checkpoint in the second line, etc.)."), information about the caller (Neuer, Fig. 14A; Col. 33, lines 36-42: "A first section 1410 may include account information of the person the agent is communicating with. This may provide account information, such as the remote party's name, account type, balance due, address, etc. Also shown is another screen section 1405 that comprises the widget with checkpoints."), a current stage of a current workflow being used for the conversation (Neuer, Fig. 13B; Col. 31, lines 27-28: “a marker may be shown associated with the current expected checkpoint or the last detected checkpoint.”), and information about the agent (Neuer, Col. 17, lines 21-25: "The CARM, in turn, is configured to inform an agent and/or administrator, often at a supervisor's computer 157 or an agent's computer 160, although other processing devices may be involved (e.g., tablets, smart phones 215, etc.)."; Here, the CARM has access to information about an agent including the computer assigned to the agent, as well as the agent's supervisor.); receiving, from the one or more models, guidance data and scoring data determined based at least partially on the behavioral and lexical features and the CRM data, wherein the guidance data includes guidance for the agent in the conversation with the caller, and the scoring data includes a rating of the conversation (Neuer, Col. 27, lines 16-26: "The positivity score may be allocated to speech conditions which represent good behavior. By allocating a negative score, bad behavior can be represented and measured. In one embodiment, these speech conditions can be linked to customer service goals. For example, the alert mapping table 1200 allocates positivity score points for providing a welcome greeting and providing a proper “wrap up” (e.g., thanking the customer and asking if there are any other questions they may assist with). On the other hand, points may be subtracted if the agent states a curse word during the call."); outputting to the CRM system, a notification comprising the guidance data and scoring data in a format associated with the CRM system (Neuer, Fig. 9; Col. 17, lines 21-32: "The CARM, in turn, is configured to inform an agent and/or administrator, often at a supervisor's computer 157 or an agent's computer 160, although other processing devices may be involved (e.g., tablets, smart phones 215, etc.). In one embodiment, the CARM processes the alert and/or event message from the RTSA system and generates the appropriate indication to the administrator. As will be seen, the CARM may map an alert message to a variety of alert indication formats, depending on various criteria. Further, the CARM may also process an event message that results in presenting and updating a checkpoint widget to the agent and/or supervisor."), wherein the guidance data comprises a suggested new stage of the current workflow for the conversation (Neuer, Fig. 13B; Col. 31, lines 40-46: “The widget shown in FIG. 13A-13B illustrates that a number of checkpoints can be expected during the call. The widget indicates in real-time which checkpoints have or have not occurred as the call progresses. This provides an easy-to-comprehend status of the call, allowing an agent or supervisor to quickly view and understand how the call is progressing.”); determining that additional audio data associated with the conversation is available since the notification was generated (Neuer, Col. 36, lines 31-38: “If the call is over in operation 1660, the process is done. Typically, the widget will be reset to its initial state for the next call. If the call is not over, then a test is made in operation 1670 to determine if another event is received. If so, the process loops back to determining whether the event is indicated as a checkpoint in operation 1640. If no event is received, but the call is still active, then the process loops back to operation 1660 to see if the call is completed.”); in response to the determination, performing behavioral and lexical analysis on the additional audio data (Neuer, Col. 36, lines 20-28: “Since not all events will necessarily be indicated as checkpoints, a test is made in operation 1640 as to whether this particular event is indicated as a checkpoint. If it is, then an update to the widget being displayed occurs in operation 1650. As explained previously, updating a widget may involve modifying the shading, color, or adding a visual element to the widget. Since the checkpoints on a widget are different, the appropriate checkpoint correlated with the event message must be updated.”; Col. 36, lines 14-19: “The process flow 1600 shown in FIG. 16 process flow presumes that at least one event is received from the RTSA system in operation 1630 (as will usually be the case). The event message reports the detection of a previously defined keyword in the speech by the agent on the voice call.”; Col. 6, lines 42-47: “Real Time Speech Analytics System (“RTSA System”)—a speech analytics system capable of providing real-time analysis of speech in a defined context. This may be employed in a contact center to detect and report on various speech conditions between an agent and a remote party that may occur during a call.”); generating feedback data related to the conversation based, at least in part, on the transmission of the notification and the behavioral and lexical analysis on the additional audio data (Neuer, Col. 34, lines 19-32: “Turning to FIG. 14B, the same screen shot 1400 is shown but with an added icon and text 1470. This comprises a warning symbol and text that reflects the speech condition that was detected for this checkpoint. This allows feedback to the agent to reinforce speech that is not desired for whatever reason. In this embodiment, the icon and text 1470 remains displayed until the agent acknowledges the alert. Further, the icon and text 1470 are only displayed once the checkpoint has been detected. In other embodiments, the alert could be displayed for a predefined time period. In other embodiments, the information may be logged into a file recording the agent's checkpoints, or a notification message could be sent to the agent's supervisor or team leader.”), and output the feedback data to the device associated with the agent in the conversation during the conversation (Neuer, Col. 33, lines 25-31: “Turing to FIG. 14A, another format of a checkpoint widget is shown. As can be appreciated, one skilled in the art could develop a number of different widget formats in light of the present specification to graphically illustrate the progress and status of various checkpoints during a call. In this embodiment, a screen shot 1400 is shown that may be presented to an agent during a call.”), wherein the guidance data comprises a suggested new stage of the current workflow for the conversation (Neuer, Fig. 13B; Col. 31, lines 40-46: “The widget shown in FIG. 13A-13B illustrates that a number of checkpoints can be expected during the call. The widget indicates in real-time which checkpoints have or have not occurred as the call progresses. This provides an easy-to-comprehend status of the call, allowing an agent or supervisor to quickly view and understand how the call is progressing.”; Col. 34, lines 4-6: “The widgets illustrated in FIGS. 13A, 13B, and 14A illustrate checkpoints that are normally expected to occur on a call.”).
McCord, Kumar, and Neuer are analogous arts because they each belong to the same field of endeavor of customer service systems. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the customer interaction and experience system of McCord, as modified by the user communication recording system of Kumar, to incorporate the teachings of Neuer to score an agent based on their interactions with a customer. The additional scoring system can be used to evaluate how well an agent is performing during customer interactions (Neuer, col. 27, lines 26-28). Evaluating the performance of employees allows a company and its employees to determine ways to improve their services. However, McCord, in view of Kumar and Neuer, fails to expressly recite applying the behavioral and lexical features and the CRM data to one or more models that classify aspects of the conversation, wherein the one or more models are structured as a plurality of processing elements comprising a directed acyclic graph executed as a parallel algorithm in which each processing element of the plurality of processing elements performs an operation as a parallel process when a set of inputs becomes available to the one or more models, and wherein the one or more models were trained to provide information on a current state of the conversation using a plurality of labeled conversations.
Sivasubramanian teaches applying the behavioral and lexical features and the CRM data to one or more models that classify aspects of the conversation (Sivasubramanian, Fig. 1, 120; [0074]: “In at least one embodiment, NLP service 132 uses artificial intelligence and/or machine learning techniques to perform sentiment analysis 120A, entity detection 120B, key phrase detection 120C, and various combinations thereof. In at least one embodiment, text-based contacts are organized by turns—for example, turns may alternate based on which party was speaking or typing on a contact. Each sentence spoken may correspond to a turn (e.g., successive turns may be from the same speaker). In at least some embodiments, each turn is analyzed separately for sentiment analysis 120A, entity detection 120B, key phrase detection 120C, and various combinations thereof. In some embodiments, for the text of a turn, sentiment analysis 120A, entity detection 120B, key phrase detection 120C, and various combinations thereof are processed in parallel by NLP service 132.”), wherein the one or more models are structured as a plurality of processing elements comprising a directed acyclic graph executed as a parallel algorithm in which each processing element of the plurality of processing elements performs an operation as a parallel process when a set of inputs becomes available to the one or more models (Sivasubramanian, Fig. 1, 114; [0068]: “Step functions workflow 114 may refer to a series of operations that are specified to run a job. The workflow may be specified in the job, either directly or indirectly. For example, the job may specify a set of capabilities of comprehend service 132 to utilize as part of the workflow. A step function workflow may include a series of steps, some of which may be executed in parallel and others which are to be executed sequentially. A workflow may specify a set of dependencies that describes how the workflow is to be executed. A workflow may be represented as a directed acyclic graph where nodes represent different steps and directed edges represent dependencies.”
[0074]: “In at least one embodiment, NLP service 132 uses artificial intelligence and/or machine learning techniques to perform sentiment analysis 120A, entity detection 120B, key phrase detection 120C, and various combinations thereof. In at least one embodiment, text-based contacts are organized by turns—for example, turns may alternate based on which party was speaking or typing on a contact. Each sentence spoken may correspond to a turn (e.g., successive turns may be from the same speaker). In at least some embodiments, each turn is analyzed separately for sentiment analysis 120A, entity detection 120B, key phrase detection 120C, and various combinations thereof. In some embodiments, for the text of a turn, sentiment analysis 120A, entity detection 120B, key phrase detection 120C, and various combinations thereof are processed in parallel by NLP service 132.”), and wherein the one or more models were trained to provide information on a current state of the conversation using a plurality of labeled conversations (Sivasubramanian, [0076]: “In at least some embodiments, a client can supply a set of training data which is used by NLP service 132 to train a neural network to recognize a custom entity.”).
McCord, Kumar, Neuer, and Sivasubramanian are analogous arts because they both belong to the same field of endeavor of customer service systems. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the customer interaction and experience system of McCord, as modified by the user communication recording system of Kumar and the agent scoring system of Neuer, to incorporate the teachings of Sivasubramanian to incorporate models running in parallel to classify aspects of a conversation. This allows multiple processes to be done at one, whenever an input is ready (Sivasubramanian, [0068]), thus allowing for more efficient processing of data.
Regarding claim 5, the rejection of claim 1 is incorporated. McCord, in view of Kumar, Neuer, and Sivasubramanian, discloses all of the elements of the current invention as stated above. McCord further discloses the method further comprising determining the behavioral and lexical features from the audio data (McCord, Col. 17, lines 5-23: "According to various aspects, when an agent (i.e., a contact center agent) interacts with a customer (typically a caller into a contact center, or a customer who has been called by the contact center), a series of measurements of semantic or other distance between the language of the agent and the language of the customer is made"; "“Distance” is used here as a term analogous to “conversational similarity”, and distance measurements may be made based on semantic distance, stylistic distance, emotional distance, acoustic difference (for example, differences in tonal range and prosody/rhythm), or other similar distance measures appropriate for measuring distance between spoken and/or written text fragments.").
Regarding claim 6, the rejection of claim 1 is incorporated. McCord, in view of Kumar, Neuer, and Sivasubramanian, discloses all of the elements of the current invention as stated above. McCord further discloses wherein determining the behavioral and lexical features comprises: identifying one or more parameters of the audio data; and utilizing the one or more parameters during the determination (McCord, Col. 17, lines 5-23: "According to various aspects, when an agent (i.e., a contact center agent) interacts with a customer (typically a caller into a contact center, or a customer who has been called by the contact center), a series of measurements of semantic or other distance between the language of the agent and the language of the customer is made"; "“Distance” is used here as a term analogous to “conversational similarity”, and distance measurements may be made based on semantic distance, stylistic distance, emotional distance, acoustic difference (for example, differences in tonal range and prosody/rhythm), or other similar distance measures appropriate for measuring distance between spoken and/or written text fragments.").
Regarding claim 7, the rejection of claim 1 is incorporated. McCord, in view of Kumar, Neuer, and Sivasubramanian, discloses all of the elements of the current invention as stated above. McCord further discloses wherein the one or more parameters include indicators of an emotional state of the second party (McCord, Col. 4, lines 21-25: "a customer interaction and experience engine which automatically gathers text-based information and performs an efficient and effective analysis and representation of the sentiment and emotion contained in the information").
Regarding claim 8, the rejection of claim 1 is incorporated. McCord, in view of Kumar, Neuer, and Sivasubramanian, discloses all of the elements of the current invention as stated above. Neuer further teaches wherein the notification comprises a rating of the performance of the agent during the conversation (Neuer, col. 27, lines 26-28: " the cumulative number of positivity points can be used to evaluate how well the agent is performing in these aspects."). The same motivation for claim 1 applies equally to claim 8.
Claim(s) 2-4, 9-11, 23, and 24 is/are rejected under 35 U.S.C. 103 as being unpatentable over McCord, in view of Kumar, Neuer, and Sivasubramanian, as applied to claims 1 and 5-8 above, and further in view of Ramachandran et al. (US Pub. No. 2022/0309413 A1 hereinafter Ramachandran).
Regarding claim 2, the rejection of claim 1 is incorporated. McCord, in view of Kumar, Neuer, and Sivasubramanian, discloses all of the elements of the current invention as stated above. Neuer further teaches wherein the one or more models comprise a behavioral model (Neuer, col. 27, lines 16-26: “In one embodiment, these speech conditions can be linked to customer service goals. For example, the alert mapping table 1200 allocates positivity score points for providing a welcome greeting and providing a proper “wrap up” (e.g., thanking the customer and asking if there are any other questions they may assist with). On the other hand, points may be subtracted if the agent states a curse word during the call.") and a call score model (Neuer, Col. 27, lines 16-26: "The positivity score may be allocated to speech conditions which represent good behavior. By allocating a negative score, bad behavior can be represented and measured.”). The same motivation for claim 1 applies equally to claim 2. However, McCord, in view of Kumar, Neuer, and Sivasubramanian, fails to expressly recite wherein the one or more models comprise a context model, a call type model, and a topic detection model.
Ramachandran teaches wherein the one or more models comprise a context model, a call type model, and a topic detection model (Ramachandran, ¶ [0024]: "The CEM 118 extracts call context 116 from the transcribed text 114. The call context includes call intent or reason, call topics or and call entities or portions of the transcribed text that help define what the call is about.").
McCord, Kumar, Neuer, Sivasubramanian, and Ramachandran are analogous arts because they both belong to the same field of endeavor of customer service systems. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the customer interaction and experience system of McCord, as modified by the user communication recording system of Kumar, the agent scoring system of Neuer, and the customer contact analysis system of Sivasubramanian, to incorporate the teachings of Ramachandran to include models for detection context, call type, and call topic. These additional models allow for an automatic computer system to identify more aspects of a customer service call, which is important because "there exists a need for improved call center computing and management systems, which can provide real-time automated guidance on a workflow to be used to an agent” (Ramachandran, ¶ [0005]).
Regarding claim 3, the rejection of claim 2 is incorporated. McCord, in view of Kumar, Neuer, Sivasubramanian, and Ramachandran, discloses all of the elements of the current invention as stated above. Ramachandran further discloses wherein the one or more models are updated based on the behavioral and lexical features and the CRM data (Ramachandran, ¶ [0028]: "In embodiments using AI/ML models, the WGM 120 is first trained using training and validation data 128 using training and validation techniques to train AI/ML models as well known in the art. The training and validation data 128 includes known best workflows that match with the customer's intent of a call, and/or workflows that match with the parameters extracted from the call context, call metadata and/or historical data, corresponding to one or more call scenarios or actual calls."). The same motivation for claim 2 applies equally to claim 3.
Regarding claim 4, the rejection of claim 1 is incorporated. McCord, in view of Kumar, Neuer, and Sivasubramanian, discloses all of the elements of the current invention as stated above. However, McCord, in view of Kumar, Neuer, and Sivasubramanian, fails to expressly recite wherein the notification comprises one or more suggestions for interacting with the caller.
Ramachandran teaches wherein the notification comprises one or more suggestions for interacting with the caller (Ramachandran, ¶ [0011]: "Embodiments of the present invention relate to a method and an apparatus for providing automated workflow guidance to an agent in a call center environment, for example, during a voice call between an agent and a customer of a business.").
McCord, Kumar, Neuer, Sivasubramanian, and Ramachandran are analogous arts because they both belong to the same field of endeavor of customer service systems. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the customer interaction and experience system of McCord, as modified by the user communication recording system of Kumar, the agent scoring system of Neuer, and the customer contact analysis system of Sivasubramanian, to incorporate the teachings of Ramachandran to include models for detection context, call type, and call topic. These additional models allow for an automatic computer system to identify more aspects of a customer service call, which is important because "there exists a need for improved call center computing and management systems, which can provide real-time automated guidance on a workflow to be used to an agent” (Ramachandran, ¶ [0005]).
Regarding claim 9, the rejection of claim 1 is incorporated. McCord, in view of Kumar, Neuer, and Sivasubramanian, discloses all of the elements of the current invention as stated above. However, McCord, in view of Kumar, Neuer, and Sivasubramanian, fails to expressly recite wherein the notification comprises an alteration of a process flow of the CRM system
Ramachandran teaches wherein the notification comprises an alteration of a process flow of the CRM system (Ramachandran, ¶ [0011]: "Embodiments of the present invention relate to a method and an apparatus for providing automated workflow guidance to an agent in a call center environment, for example, during a voice call between an agent and a customer of a business.").
McCord, Kumar, Neuer, Sivasubramanian, and Ramachandran are analogous arts because they both belong to the same field of endeavor of customer service systems. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the customer interaction and experience system of McCord, as modified by the user communication recording system of Kumar, the agent scoring system of Neuer, and the customer contact analysis system of Sivasubramanian, to incorporate the teachings of Ramachandran to include models for detection context, call type, and call topic. These additional models allow for an automatic computer system to identify more aspects of a customer service call, which is important because "there exists a need for improved call center computing and management systems, which can provide real-time automated guidance on a workflow to be used to an agent” (Ramachandran, ¶ [0005]).
Regarding claim 10, the rejection of claim 1 is incorporated. McCord, in view of Kumar, Neuer, and Sivasubramanian, discloses all of the elements of the current invention as stated above. Neuer further teaches wherein the scoring data is utilized by the CRM system during the conversation to affect the conversation (Neuer, Fig. 9; Col 24, lines 49-62: “the pop-up window 875 may include text portions 910, 915 that reflect the text with prior alerts. The agent may choose to read and mark these as having been reviewed. Separate controls may be presented 930 allowing the agent to review all of the alerts for the agent or only those unread. In addition, another portion 905 of the pop-up window includes a positivity score 920 and a compliance score 925. These scores provide a summary of the current standing on the agent's performance relative to customer service (e.g., “positivity”) and compliance. These scores are assigned to each detected RTSA topic and allow a weighting of the relative importance of the detected speech conditions. In other embodiments, the scores may be defined for a different category or purpose."). The same motivation for claim 1 applies equally to claim 10. However, McCord, in view of Kumar, Neuer, and Sivasubramanian, fails to expressly recite wherein the guidance data is utilized by the CRM system during the conversation to affect the conversation.
Ramachandran teaches wherein the guidance data is utilized by the CRM system during the conversation to affect the conversation (Ramachandran, ¶ [0011]: "Embodiments of the present invention relate to a method and an apparatus for providing automated workflow guidance to an agent in a call center environment, for example, during a voice call between an agent and a customer of a business.").
McCord, Kumar, Neuer, Sivasubramanian, and Ramachandran are analogous arts because they both belong to the same field of endeavor of customer service systems. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the customer interaction and experience system of McCord, as modified by the user communication recording system of Kumar, the agent scoring system of Neuer, and the customer contact analysis system of Sivasubramanian, to incorporate the teachings of Ramachandran to include models for detection context, call type, and call topic. These additional models allow for an automatic computer system to identify more aspects of a customer service call, which is important because "there exists a need for improved call center computing and management systems, which can provide real-time automated guidance on a workflow to be used to an agent” (Ramachandran, ¶ [0005]).
Regarding claim 11, the rejection of claim 1 is incorporated. McCord, in view of Kumar, Neuer, and Sivasubramanian, discloses all of the elements of the current invention as stated above. Neuer further teaches wherein the scoring data is utilized by the CRM system to affect a subsequent communication session (Neuer, Fig. 9; Col 24, lines 49-62: “the pop-up window 875 may include text portions 910, 915 that reflect the text with prior alerts. The agent may choose to read and mark these as having been reviewed. Separate controls may be presented 930 allowing the agent to review all of the alerts for the agent or only those unread. In addition, another portion 905 of the pop-up window includes a positivity score 920 and a compliance score 925. These scores provide a summary of the current standing on the agent's performance relative to customer service (e.g., “positivity”) and compliance. These scores are assigned to each detected RTSA topic and allow a weighting of the relative importance of the detected speech conditions. In other embodiments, the scores may be defined for a different category or purpose."; Col. 18, line 60 - Col. 19, line 5: "The CARM may also provide the meta-data related to the call to the file store 190 via signaling 235. For example, for each call stored in the file store 190, the CARM may provide meta-data relating to timing information of checkpoints that occurred during the call. The timing information, as will be seen, may be defined in various formats. One format may be an offset time from the beginning of each call for each checkpoint. For example, the meta-data may have an offset time #1 for the first checkpoint, an offset time #2 for the second checkpoint, etc. The file store 190 may store this information in association with the call recording, so that the call recording and the meta-data can be retrieved at a later time for reviewing the checkpoint widget."). The same motivation for claim 1 applies equally to claim 11. However, McCord, in view of Kumar, Neuer, and Sivasubramanian, fails to expressly recite wherein the guidance data is utilized by the CRM system to affect a subsequent communication session.
Ramachandran teaches wherein the guidance data is utilized by the CRM system to affect a subsequent communication session (Ramachandran, ¶ [0028]: "In embodiments using AI/ML models, the WGM 120 is first trained using training and validation data 128 using training and validation techniques to train AI/ML models as well known in the art. The training and validation data 128 includes known best workflows that match with the customer's intent of a call, and/or workflows that match with the parameters extracted from the call context, call metadata and/or historical data, corresponding to one or more call scenarios or actual calls.").
McCord, Kumar, Neuer, Sivasubramanian, and Ramachandran are analogous arts because they both belong to the same field of endeavor of customer service systems. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the customer interaction and experience system of McCord, as modified by the user communication recording system of Kumar, the agent scoring system of Neuer, and the customer contact analysis system of Sivasubramanian, to incorporate the teachings of Ramachandran to include models for detection context, call type, and call topic. These additional models allow for an automatic computer system to identify more aspects of a customer service call, which is important because "there exists a need for improved call center computing and management systems, which can provide real-time automated guidance on a workflow to be used to an agent” (Ramachandran, ¶ [0005]).
Regarding claim 23, the rejection of claim 1 is incorporated. McCord, in view of Kumar, Neuer, and Sivasubramanian, discloses all of the elements of the current invention as stated above. However, McCord, in view of Kumar, Neuer, and Sivasubramanian, fails to expressly recite wherein the information about the agent comprises a conversation history of the agent.
Ramachandran teaches wherein the information about the agent comprises a conversation history of the agent (Ramachandran, [0011]: "Workflow guidance is provided based on one or more of call context extracted from the transcribed text of the conversation, call metadata and historical data."; [0012]: " Call metadata includes information related to the call, from which information such as customer profile, history of calls including previously suggested workflow, previous action adopted by the customer and results obtained therefrom and the like, are extracted.").
McCord, Kumar, Neuer, Sivasubramanian, and Ramachandran are analogous arts because they both belong to the same field of endeavor of customer service systems. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the customer interaction and experience system of McCord, as modified by the user communication recording system of Kumar, the agent scoring system of Neuer, and the customer contact analysis system of Sivasubramanian, to incorporate the teachings of Ramachandran to include models for detection context, call type, and call topic. These additional models allow for an automatic computer system to identify more aspects of a customer service call, which is important because "there exists a need for improved call center computing and management systems, which can provide real-time automated guidance on a workflow to be used to an agent” (Ramachandran, ¶ [0005]).
Regarding claim 24, the rejection of claim 1 is incorporated. McCord, in view of Kumar, Neuer, and Sivasubramanian, discloses all of the elements of the current invention as stated above. However, McCord, in view of Kumar, Neuer, and Sivasubramanian, fails to expressly recite wherein the information about the agent comprises one or more performance reviews of the agent.
Ramachandran teaches wherein the information about the agent comprises one or more performance reviews of the agent (Ramachandran, [0012]: "Historical data typically includes customer satisfaction history, call resolution history, feedback history and other speech analytics derived from previous calls of the customer.").
McCord, Kumar, Neuer, Sivasubramanian, and Ramachandran are analogous arts because they both belong to the same field of endeavor of customer service systems. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the customer interaction and experience system of McCord, as modified by the user communication recording system of Kumar, the agent scoring system of Neuer, and the customer contact analysis system of Sivasubramanian, to incorporate the teachings of Ramachandran to include models for detection context, call type, and call topic. These additional models allow for an automatic computer system to identify more aspects of a customer service call, which is important because "there exists a need for improved call center computing and management systems, which can provide real-time automated guidance on a workflow to be used to an agent” (Ramachandran, ¶ [0005]).
Claim(s) 25 is/are rejected under 35 U.S.C. 103 as being unpatentable over McCord, in view of Kumar, Neuer, and Sivasubramanian, as applied to claims 1 and 5-8 above, and further in view of Wang et al. (CN Pat. No. 110444223 A hereinafter Wang).
Regarding claim 25, the rejection of claim 1 is incorporated. McCord, in view of Kumar, Neuer, and Sivasubramanian, discloses all of the elements of the current invention as stated above. However, McCord, in view of Kumar, Neuer, and Sivasubramanian, fails to expressly recite wherein the behavioral and lexical features are determined by grouping acoustic spectral measurements in a time interval of individual words and mapping the acoustic spectral measurements to a one-dimensional vector representation to produce word-aligned, non-verbal embeddings.
Wang teaches wherein the behavioral and lexical features are determined by grouping acoustic spectral measurements in a time interval of individual words and mapping the acoustic spectral measurements to a one-dimensional vector representation to produce word-aligned, non-verbal embeddings (Wang, Pg. 5, final paragraph: “In this embodiment, when carrying out feature extracting MFCC feature vector set from the to-be-identified voice (name of MFCC is Mel-scale Frequency Cepstral Coefficients, Mel reverse spectrum coefficient), can preset window length of said characteristic window is 25ms, the preset moving distance of the feature window is 10ms. from the window length 25ms, the moving distance in the window of 10ms extracting 13 dimensional MFCC feature, and then average the sameness, each word to obtain a one-dimensional column vector of 13 * 1 so as to compose the MFCC feature vector set. by extracting the MFCC characteristic vector corresponding with the voice to be identified, effective physical information (the speech spectral envelope and details) to a coding operation to obtain the set of feature vectors.”).
McCord, Kumar, Neuer, Sivasubramanian, and Wang are analogous arts because they each belong to the same field of endeavor of customer service systems. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the customer interaction and experience system of McCord, as modified by the user communication recording system of Kumar, the agent scoring system of Neuer, and the customer contact analysis system of Sivasubramanian, to incorporate the teachings of Wang to create a per-word group of spectral measurements in the form of one-dimensional vectors. This ensures that the system can effectively collect physical information about the speech for further processing (Wang, pg. 5). As such, the system is able to have spectral data in a standardized form, making for easier future processing.
Claim(s) 26 is/are rejected under 35 U.S.C. 103 as being unpatentable over McCord, in view of Kumar, Neuer, and Sivasubramanian, as applied to claims 1 and 5-8 above, and further in view of Kwatra, Shikhar (US Pat. Pub. No. 2020/0043355 A1 hereinafter Kwatra).
Regarding claim 26, the rejection of claim 1 is incorporated. McCord, in view of Kumar, Neuer, and Sivasubramanian, discloses all of the elements of the current invention as stated above. However, McCord, in view of Kumar, Neuer, and Sivasubramanian, fails to expressly recite wherein outputting the notification to the CRM system comprises continuously polling, by a customer relationship management (CRM) data processor, for a guidance notification from a guidance integration processor.
Kwatra teaches wherein outputting the notification to the CRM system comprises continuously polling, by a customer relationship management (CRM) data processor, for a guidance notification from a guidance integration processor (Kwatra, [0059]: “One or more mistakes committed by the users can be stored in the feature matching operation in order to trigger or activate the mistaken steps before the actual steps so that the user keeps track of the correct instructions or steps to be executed. That is, if the user commits certain mistakes (e.g., skipping some steps or proceeding incorrectly with repairing something), then feature matching system, which is a part of monitoring user's activities correlating with ideal activity to be performed while fixing or repairing something (e.g., continuous software polling being performed by monitoring the user's activity and the preferred way of doing the activity), may calculate a threshold deviation from ideal way of performing the activity or completing a certain task and then corrective actions can be provided to the user by interjecting at a certain point and guiding the user with a modified set of instructions.”).
McCord, Kumar, Neuer, Sivasubramanian, and Kwatra are analogous arts because they each belong to the same field of endeavor of customer service systems. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the customer interaction and experience system of McCord, as modified by the user communication recording system of Kumar, the agent scoring system of Neuer, and the customer contact analysis system of Sivasubramanian, to incorporate the teachings of Kwatra to continuously poll for guidance notifications. Continuously polling for guidance allows the system to quickly and automatically respond to a user’s errors (Kwatra, [0059]). This ensures that the system can effectively provide guidance to the user.
Claim(s) 12, 13, 15, and 16 is/are rejected under 35 U.S.C. 103 as being unpatentable over McCord, in view of Kumar, Griffin, Neuer, and Sivasubramanian.
Regarding claim 12, McCord discloses a system for outputting feedback data to a selected device in a call center (McCord, Col. 11, lines 61-66: “FIG. 3 is a block diagram illustrating an exemplary system architecture for a contact center utilizing a self-learning interaction optimizer (SLIO) 300 comprising a reinforcement learning server 310 and an optimization server 320 (both shown below in FIG. 4), according to one aspect.”), comprising: a memory configured to store representations of data in an electronic form; and a processor operatively coupled to the memory, the processor configured to access the data and process the data to (McCord, Col. 19, lines 42-50: “CPU 12 may include one or more processors 13 such as, for example, a processor from one of the Intel, ARM, Qualcomm, and AMD families of microprocessors. In some aspects, processors 13 may include specially designed hardware such as application-specific integrated circuits (ASICs), electrically erasable programmable read-only memories (EEPROMs), field-programmable gate arrays (FPGAs), and so forth, for controlling operations of computing device 10.”): access audio data, wherein the audio data is associated with a conversation between an agent of a call center and a caller, perform behavioral and lexical analysis on the audio data, extract features based on the behavioral and lexical analysis (McCord, Col. 17, lines 5-23: "According to various aspects, when an agent (i.e., a contact center agent) interacts with a customer (typically a caller into a contact center, or a customer who has been called by the contact center), a series of measurements of semantic or other distance between the language of the agent and the language of the customer is made"; "“Distance” is used here as a term analogous to “conversational similarity”, and distance measurements may be made based on semantic distance, stylistic distance, emotional distance, acoustic difference (for example, differences in tonal range and prosody/rhythm), or other similar distance measures appropriate for measuring distance between spoken and/or written text fragments."), apply machine learning on the extracted features using one or more models, wherein the one or more models were trained using a plurality of labeled conversations (McCord, Col. 14, lines 31-34: "The sentiment and emotion analyzer 1413 analyzes the incoming text for sentiment and emotional content and forwards the analysis to the scoring and display engine 1414."), generate a notification based at least in part on the machine learning (McCord, Col. 14, lines 34-40: "The scoring and display engine 1414 cumulatively assigns scores for sentiment and emotion to products and services of interest to the business based on the analysis from the sentiment and emotion analyzer 1413, and displays the resulting information in textual, tabular, or graphical form for use by decision-makers at the business."). However, McCord fails to expressly recite collect information currently being displayed on a device associated with the agent; wherein the one or more models are structured as a plurality of processing elements comprising a directed acyclic graph executed as a parallel algorithm in which each processing element of the plurality of processing elements performs an operation as a parallel process when a set of inputs becomes available to the one or more models; determine that the notification includes customer relationship management data, wherein, upon the determination that the notification includes customer relationship management data, transmitting the notification to a customer relationship management integration device, wherein the customer relationship management data comprises input from the agent, the information displayed on the device associated with the agent, management flow data associated with the conversation, information about the caller, a current stage of a current workflow being used for the conversation, and information about the agent, determine that additional audio data associated with the conversation is available since the notification was generated, in response to the determination, perform behavioral and lexical analysis on the additional audio data, generate feedback data related to the conversation based, at least in part, on the transmission of the notification and the behavioral and lexical analysis on the additional audio data, and output the feedback data to the device associated with the agent in the conversation during the conversation, wherein the guidance data comprises a suggested new stage of the current workflow for the conversation.
Kumar teaches collect information currently being displayed on a device associated with the agent; and wherein the customer relationship management data comprises the information displayed on the device associated with the agent (Kumar, [0027]: "The recording of the communications session may be performed in addition, or as an alternative, to routing the communications session to human agents for handling. The recording may be performed in one or both of two distinct ways: (1) the customer's interaction with the human agent (e.g., an audio recording of the speech-based conversation between the customer and the agent during a voice call), or the interactive response system, or both may be recorded, and/or (2) the computer screens presented to the human agent by the agent's computer during the agent's interaction with the customer may be recorded or captured (often referred to as “screen captures”).").
McCord and Kumar are analogous arts because they both belong to the same field of endeavor
of customer service systems. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the customer interaction and experience system of McCord to incorporate the teachings of Kumar to capture information from the screen of an agent. Recording what is being displayed on an agent’s screen can help monitor the agent’s effectiveness (Kumar, [0028]). This ensures that any issues with an agent’s effectiveness can be properly addressed. However, McCord, in view of Kumar, fails to expressly recite wherein the one or more models are structured as a plurality of processing elements comprising a directed acyclic graph executed as a parallel algorithm in which each processing element of the plurality of processing elements performs an operation as a parallel process when a set of inputs becomes available to the one or more models; determine that the notification includes customer relationship management data, wherein, upon the determination that the notification includes customer relationship management data, transmitting the notification to a customer relationship management integration device, wherein the customer relationship management data comprises input from the agent, management flow data associated with the conversation, information about the caller, a current stage of a current workflow being used for the conversation, and information about the agent, determine that additional audio data associated with the conversation is available since the notification was generated, in response to the determination, perform behavioral and lexical analysis on the additional audio data, generate feedback data related to the conversation based, at least in part, on the transmission of the notification and the behavioral and lexical analysis on the additional audio data, and output the feedback data to the device associated with the agent in the conversation during the conversation, wherein the guidance data comprises a suggested new stage of the current workflow for the conversation.
Griffin teaches determine that the notification includes customer relationship management data (Griffin, Fig. 2, ¶ [0032]: "It is then determined whether the query pertains to data available on a public cloud resource, or the query pertains to data available on a private cloud resource (step 212)"), wherein, upon the determination that the notification includes customer relationship management data, transmitting the notification to a customer relationship management integration device (Griffin, Fig. 2, ¶ [0033]: " if it is determined that the query pertains to the data available on the private cloud resource, the query is interpreted by the using an enterprise specific model trained on at least one machine learning technique on data from the private cloud (step 216).").
McCord, Kumar, and Griffin are analogous arts because they each belong to the same field of endeavor of data processing. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the customer interaction and experience system of McCord, as modified by the user communication recording system of Kumar, to incorporate the teachings of Griffin to identify the type of data in a transfer, and send that data to an appropriate system. Sending CRM data, including user data, to a different system then general data allows for additional security and protection of sensitive data (Griffin, ¶ [0003]). It is important to protect a user’s sensitive data so that no one gains unauthorized access to sensitive information. However, McCord, in view of Kumar and Griffin, fails to expressly recite wherein the one or more models are structured as a plurality of processing elements comprising a directed acyclic graph executed as a parallel algorithm in which each processing element of the plurality of processing elements performs an operation as a parallel process when a set of inputs becomes available to the one or more models; wherein the customer relationship management data comprises input from the agent, management flow data associated with the conversation, information about the caller, a current stage of a current workflow being used for the conversation, and information about the agent, determine that additional audio data associated with the conversation is available since the notification was generated, in response to the determination, perform behavioral and lexical analysis on the additional audio data, generate feedback data related to the conversation based, at least in part, on the transmission of the notification and the behavioral and lexical analysis on the additional audio data, and output the feedback data to the device associated with the agent in the conversation during the conversation, wherein the guidance data comprises a suggested new stage of the current workflow for the conversation.
Neuer teaches wherein the customer relationship management data comprises input from the agent (Neuer, Col. 19, lines 2-5: "The file store 190 may store this information in association with the call recording, so that the call recording and the meta-data can be retrieved at a later time for reviewing the checkpoint widget."; Here, the call recording includes speech input from the agent.), management flow data associated with the conversation (Neuer, Fig. 14A; Col. 33, lines 42-48: "In this embodiment, the checkpoints 1410, 1420, 1430, 1440, 1450 are arranged as a series of checkboxes along with text in a list format. There is an implied time-line reflecting the passage of time in a downward manner (e.g., the first checkpoint associated with the first line is expected to occur before the second checkpoint in the second line, etc.)."), information about the caller (Neuer, Fig. 14A; Col. 33, lines 36-42: "A first section 1410 may include account information of the person the agent is communicating with. This may provide account information, such as the remote party's name, account type, balance due, address, etc. Also shown is another screen section 1405 that comprises the widget with checkpoints."), a current stage of a current workflow being used for the conversation (Neuer, Fig. 13B; Col. 31, lines 27-28: “a marker may be shown associated with the current expected checkpoint or the last detected checkpoint.”), and information about the agent (Neuer, Col. 17, lines 21-25: "The CARM, in turn, is configured to inform an agent and/or administrator, often at a supervisor's computer 157 or an agent's computer 160, although other processing devices may be involved (e.g., tablets, smart phones 215, etc.)."; Here, the CARM has access to information about an agent including the computer assigned to the agent, as well as the agent's supervisor.), determine that additional audio data associated with the conversation is available since the notification was generated (Neuer, Col. 36, lines 31-38: “If the call is over in operation 1660, the process is done. Typically, the widget will be reset to its initial state for the next call. If the call is not over, then a test is made in operation 1670 to determine if another event is received. If so, the process loops back to determining whether the event is indicated as a checkpoint in operation 1640. If no event is received, but the call is still active, then the process loops back to operation 1660 to see if the call is completed.”), in response to the determination, perform behavioral and lexical analysis on the additional audio data (Neuer, Col. 36, lines 20-28: “Since not all events will necessarily be indicated as checkpoints, a test is made in operation 1640 as to whether this particular event is indicated as a checkpoint. If it is, then an update to the widget being displayed occurs in operation 1650. As explained previously, updating a widget may involve modifying the shading, color, or adding a visual element to the widget. Since the checkpoints on a widget are different, the appropriate checkpoint correlated with the event message must be updated.”; Col. 36, lines 14-19: “The process flow 1600 shown in FIG. 16 process flow presumes that at least one event is received from the RTSA system in operation 1630 (as will usually be the case). The event message reports the detection of a previously defined keyword in the speech by the agent on the voice call.”; Col. 6, lines 42-47: “Real Time Speech Analytics System (“RTSA System”)—a speech analytics system capable of providing real-time analysis of speech in a defined context. This may be employed in a contact center to detect and report on various speech conditions between an agent and a remote party that may occur during a call.”), generate feedback data related to the conversation based, at least in part, on the transmission of the notification and the behavioral and lexical analysis on the additional audio data (Neuer, Col. 34, lines 19-32: “Turning to FIG. 14B, the same screen shot 1400 is shown but with an added icon and text 1470. This comprises a warning symbol and text that reflects the speech condition that was detected for this checkpoint. This allows feedback to the agent to reinforce speech that is not desired for whatever reason. In this embodiment, the icon and text 1470 remains displayed until the agent acknowledges the alert. Further, the icon and text 1470 are only displayed once the checkpoint has been detected. In other embodiments, the alert could be displayed for a predefined time period. In other embodiments, the information may be logged into a file recording the agent's checkpoints, or a notification message could be sent to the agent's supervisor or team leader.”), and output the feedback data to the device associated with the agent in the conversation during the conversation (Neuer, Col. 33, lines 25-31: “Turing to FIG. 14A, another format of a checkpoint widget is shown. As can be appreciated, one skilled in the art could develop a number of different widget formats in light of the present specification to graphically illustrate the progress and status of various checkpoints during a call. In this embodiment, a screen shot 1400 is shown that may be presented to an agent during a call.”), wherein the guidance data comprises a suggested new stage of the current workflow for the conversation (Neuer, Fig. 13B; Col. 31, lines 40-46: “The widget shown in FIG. 13A-13B illustrates that a number of checkpoints can be expected during the call. The widget indicates in real-time which checkpoints have or have not occurred as the call progresses. This provides an easy-to-comprehend status of the call, allowing an agent or supervisor to quickly view and understand how the call is progressing.”; Col. 34, lines 4-6: “The widgets illustrated in FIGS. 13A, 13B, and 14A illustrate checkpoints that are normally expected to occur on a call.”).
McCord, Kumar, Griffin, and Neuer are analogous arts because they all belong to the same field of endeavor of customer service systems. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the customer interaction and experience system of McCord, as modified by the user communication recording system of Kumar and the data privacy system of Griffin, to incorporate the teachings of Neuer to score an agent based on their interactions with a customer. The additional scoring system can be used to evaluate how well an agent is performing during customer interactions (Neuer, col. 27, lines 26-28). Evaluating the performance of employees allows a company and its employees to determine ways to improve their services. However, McCord, in view of Kumar, Griffin, and Neuer, fails to expressly recite wherein the one or more models are structured as a plurality of processing elements comprising a directed acyclic graph executed as a parallel algorithm in which each processing element of the plurality of processing elements performs an operation as a parallel process when a set of inputs becomes available to the one or more models.
Sivasubramanian teaches wherein the one or more models are structured as a plurality of processing elements comprising a directed acyclic graph executed as a parallel algorithm in which each processing element of the plurality of processing elements performs an operation as a parallel process when a set of inputs becomes available to the one or more models (Sivasubramanian, Fig. 1, 114; [0068]: “Step functions workflow 114 may refer to a series of operations that are specified to run a job. The workflow may be specified in the job, either directly or indirectly. For example, the job may specify a set of capabilities of comprehend service 132 to utilize as part of the workflow. A step function workflow may include a series of steps, some of which may be executed in parallel and others which are to be executed sequentially. A workflow may specify a set of dependencies that describes how the workflow is to be executed. A workflow may be represented as a directed acyclic graph where nodes represent different steps and directed edges represent dependencies.”).
McCord, Kumar, Griffin, Neuer, and Sivasubramanian are analogous arts because they both belong to the same field of endeavor of customer service systems. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the customer interaction and experience system of McCord, as modified by the user communication recording system of Kumar, the data privacy system of Griffin, and the agent scoring system of Neuer, to incorporate the teachings of Sivasubramanian to incorporate models running in parallel to classify aspects of a conversation. This allows multiple processes to be done at one, whenever an input is ready (Sivasubramanian, [0068]), thus allowing for more efficient processing of data.
Regarding claim 13, the rejection of claim 12 is incorporated. McCord, in view of Kumar, Griffin, Neuer, and Sivasubramanian, discloses all of the elements of the current invention as stated above. Griffin further teaches wherein, upon determination that the notification does not include customer relationship management data, transmit the notification to a guidance integration device (Griffin, Fig. 2, ¶ [0033]: "If it is determined that the query pertains to data available on a public cloud resource, the query is interpreted by using a general model trained on at least one machine learning technique on data from the public cloud (step 214)."). The same motivation for claim 12 applies equally to claim 13.
Regarding claim 15, the rejection of claim 12 is incorporated. McCord, in view of Kumar, Griffin, Neuer, and Sivasubramanian, discloses all of the elements of the current invention as stated above. McCord further discloses wherein the processor is further configured to: identify one or more parameters of the audio data (McCord, Col. 17, lines 5-23: "According to various aspects, when an agent (i.e., a contact center agent) interacts with a customer (typically a caller into a contact center, or a customer who has been called by the contact center), a series of measurements of semantic or other distance between the language of the agent and the language of the customer is made"; "“Distance” is used here as a term analogous to “conversational similarity”, and distance measurements may be made based on semantic distance, stylistic distance, emotional distance, acoustic difference (for example, differences in tonal range and prosody/rhythm), or other similar distance measures appropriate for measuring distance between spoken and/or written text fragments."); and utilize the one or more parameters during the performing behavioral and lexical analysis on the audio data (McCord, Col. 14, lines 31-34: "The sentiment and emotion analyzer 1413 analyzes the incoming text for sentiment and emotional content and forwards the analysis to the scoring and display engine 1414.").
Regarding claim 16, the rejection of claim 12 is incorporated. McCord, in view of Kumar, Griffin, Neuer, and Sivasubramanian, discloses all of the elements of the current invention as stated above. McCord further discloses wherein the parameters include indicators of an emotional state of the caller (McCord, Col. 14, lines 31-34: "The sentiment and emotion analyzer 1413 analyzes the incoming text for sentiment and emotional content and forwards the analysis to the scoring and display engine 1414.").
Claim(s) 19 is/are rejected under 35 U.S.C. 103 as being unpatentable over McCord, in view of Kumar, Neuer, Griffin, and Sivasubramanian.
Regarding claim 19, McCord discloses a method for providing feedback related to a call session in a call center comprising (McCord, Col. 11, lines 61-66: “FIG. 3 is a block diagram illustrating an exemplary system architecture for a contact center utilizing a self-learning interaction optimizer (SLIO) 300 comprising a reinforcement learning server 310 and an optimization server 320 (both shown below in FIG. 4), according to one aspect.”): accessing audio data that includes behavioral information and lexical information, wherein the audio data is associated with a conversation between a first party of a call center and a second party (McCord, Col. 2, lines 40-45: "Direct customer communications with the business in text form (e.g., email and live chats) are sent directly to the sentiment and emotion analyzer. Direct customer communications with the business that contain audio data (e.g., phone calls and video conferences) are sent to the automated speech recognition engine."); extracting the behavioral information and lexical information from the audio data (McCord, Col. 17, lines 5-23: "According to various aspects, when an agent (i.e., a contact center agent) interacts with a customer (typically a caller into a contact center, or a customer who has been called by the contact center), a series of measurements of semantic or other distance between the language of the agent and the language of the customer is made"; "“Distance” is used here as a term analogous to “conversational similarity”, and distance measurements may be made based on semantic distance, stylistic distance, emotional distance, acoustic difference (for example, differences in tonal range and prosody/rhythm), or other similar distance measures appropriate for measuring distance between spoken and/or written text fragments."); accessing customer relationship management analysis signals in real-time (McCord, Col. 2, lines 40-45: "Direct customer communications with the business in text form (e.g., email and live chats) are sent directly to the sentiment and emotion analyzer. Direct customer communications with the business that contain audio data (e.g., phone calls and video conferences) are sent to the automated speech recognition engine."). However, McCord fails to expressly recite collecting, from a device associated with the first party, information currently being displayed on the device; wherein the customer relationship management analysis signals comprises input from the first party, the information displayed on the device associated with the first party, management flow data associated with the conversation, information about the second party, a current stage of a current workflow being used for the conversation, and information about the first party; combining the customer relationship management analysis signals, behavioral information, and lexical information to produce guidance and scoring signals using one or more models, wherein the one or more models are structured as a plurality of processing elements comprising a directed acyclic graph executed as a parallel algorithm in which each processing element of the plurality of processing elements performs an operation as a parallel process when a set of inputs becomes available to the one or more models, and wherein the one or more models were trained using a plurality of labeled conversations; generating a notification based on the guidance and scoring signals; determine that the notification includes customer relationship management data; in response to the determination that the notification includes customer relationship management data, transmitting the notification to a customer relationship management device; determine that additional audio data associated with the conversation is available since the notification was generated; in response to the determination, perform behavioral and lexical analysis on the additional audio data; and outputting guidance data to the device associated with the first party in the conversation during the conversation to provide the first party feedback related to the conversation, wherein the guidance data comprises a suggested new stage of the current workflow for the conversation, wherein the guidance data is based, at least in part, on the transmission of the notification and the and the behavioral and lexical analysis on the additional audio data.
Kumar teaches collecting, from a device associated with the first party, information currently being displayed on the device; and wherein the customer relationship management analysis signals comprises the information displayed on the device associated with the first party (Kumar, [0027]: "The recording of the communications session may be performed in addition, or as an alternative, to routing the communications session to human agents for handling. The recording may be performed in one or both of two distinct ways: (1) the customer's interaction with the human agent (e.g., an audio recording of the speech-based conversation between the customer and the agent during a voice call), or the interactive response system, or both may be recorded, and/or (2) the computer screens presented to the human agent by the agent's computer during the agent's interaction with the customer may be recorded or captured (often referred to as “screen captures”).").
McCord and Kumar are analogous arts because they both belong to the same field of endeavor
of customer service systems. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the customer interaction and experience system of McCord to incorporate the teachings of Kumar to capture information from the screen of an agent. Recording what is being displayed on an agent’s screen can help monitor the agent’s effectiveness (Kumar, [0028]). This ensures that any issues with an agent’s effectiveness can be properly addressed. However, McCord, in view of Kumar, fails to expressly recite wherein the customer relationship management analysis signals comprises input from the first party, management flow data associated with the conversation, information about the second party, a current stage of a current workflow being used for the conversation, and information about the first party; combining the customer relationship management analysis signals, behavioral information, and lexical information to produce guidance and scoring signals using one or more models, wherein the one or more models are structured as a plurality of processing elements comprising a directed acyclic graph executed as a parallel algorithm in which each processing element of the plurality of processing elements performs an operation as a parallel process when a set of inputs becomes available to the one or more models, and wherein the one or more models were trained using a plurality of labeled conversations; generating a notification based on the guidance and scoring signals; determine that the notification includes customer relationship management data; in response to the determination that the notification includes customer relationship management data, transmitting the notification to a customer relationship management device; determine that additional audio data associated with the conversation is available since the notification was generated; in response to the determination, perform behavioral and lexical analysis on the additional audio data; and outputting guidance data to the device associated with the first party in the conversation during the conversation to provide the first party feedback related to the conversation, wherein the guidance data comprises a suggested new stage of the current workflow for the conversation, wherein the guidance data is based, at least in part, on the transmission of the notification and the and the behavioral and lexical analysis on the additional audio data.
Neuer teaches wherein the customer relationship management analysis signals comprises input from the first party (Neuer, Col. 19, lines 2-5: "The file store 190 may store this information in association with the call recording, so that the call recording and the meta-data can be retrieved at a later time for reviewing the checkpoint widget."; Here, the call recording includes speech input from the agent.), management flow data associated with the conversation (Neuer, Fig. 14A; Col. 33, lines 42-48: "In this embodiment, the checkpoints 1410, 1420, 1430, 1440, 1450 are arranged as a series of checkboxes along with text in a list format. There is an implied time-line reflecting the passage of time in a downward manner (e.g., the first checkpoint associated with the first line is expected to occur before the second checkpoint in the second line, etc.)."), information about the second party (Neuer, Fig. 14A; Col. 33, lines 36-42: "A first section 1410 may include account information of the person the agent is communicating with. This may provide account information, such as the remote party's name, account type, balance due, address, etc. Also shown is another screen section 1405 that comprises the widget with checkpoints."), a current stage of a current workflow being used for the conversation (Neuer, Fig. 13B; Col. 31, lines 27-28: “a marker may be shown associated with the current expected checkpoint or the last detected checkpoint.”), and information about the first party (Neuer, Col. 17, lines 21-25: "The CARM, in turn, is configured to inform an agent and/or administrator, often at a supervisor's computer 157 or an agent's computer 160, although other processing devices may be involved (e.g., tablets, smart phones 215, etc.)."; Here, the CARM has access to information about an agent including the computer assigned to the agent, as well as the agent's supervisor.); combining the customer relationship management analysis signals, behavioral information, and lexical information to produce guidance and scoring signals using one or more models, wherein the one or more models were trained using a plurality of labeled conversations (Neuer, Col. 27, lines 16-26: "The positivity score may be allocated to speech conditions which represent good behavior. By allocating a negative score, bad behavior can be represented and measured. In one embodiment, these speech conditions can be linked to customer service goals. For example, the alert mapping table 1200 allocates positivity score points for providing a welcome greeting and providing a proper “wrap up” (e.g., thanking the customer and asking if there are any other questions they may assist with). On the other hand, points may be subtracted if the agent states a curse word during the call."); generating a notification based on the guidance and scoring signals (Neuer, Fig. 9; Col. 17, lines 21-32: "The CARM, in turn, is configured to inform an agent and/or administrator, often at a supervisor's computer 157 or an agent's computer 160, although other processing devices may be involved (e.g., tablets, smart phones 215, etc.). In one embodiment, the CARM processes the alert and/or event message from the RTSA system and generates the appropriate indication to the administrator. As will be seen, the CARM may map an alert message to a variety of alert indication formats, depending on various criteria. Further, the CARM may also process an event message that results in presenting and updating a checkpoint widget to the agent and/or supervisor."); determine that additional audio data associated with the conversation is available since the notification was generated (Neuer, Col. 36, lines 31-38: “If the call is over in operation 1660, the process is done. Typically, the widget will be reset to its initial state for the next call. If the call is not over, then a test is made in operation 1670 to determine if another event is received. If so, the process loops back to determining whether the event is indicated as a checkpoint in operation 1640. If no event is received, but the call is still active, then the process loops back to operation 1660 to see if the call is completed.”); in response to the determination, perform behavioral and lexical analysis on the additional audio data (Neuer, Col. 36, lines 20-28: “Since not all events will necessarily be indicated as checkpoints, a test is made in operation 1640 as to whether this particular event is indicated as a checkpoint. If it is, then an update to the widget being displayed occurs in operation 1650. As explained previously, updating a widget may involve modifying the shading, color, or adding a visual element to the widget. Since the checkpoints on a widget are different, the appropriate checkpoint correlated with the event message must be updated.”; Col. 36, lines 14-19: “The process flow 1600 shown in FIG. 16 process flow presumes that at least one event is received from the RTSA system in operation 1630 (as will usually be the case). The event message reports the detection of a previously defined keyword in the speech by the agent on the voice call.”; Col. 6, lines 42-47: “Real Time Speech Analytics System (“RTSA System”)—a speech analytics system capable of providing real-time analysis of speech in a defined context. This may be employed in a contact center to detect and report on various speech conditions between an agent and a remote party that may occur during a call.”); and outputting guidance data to the device associated with the first party in the conversation during the conversation to provide the first party feedback related to the conversation (Neuer, Col. 33, lines 25-31: “Turing to FIG. 14A, another format of a checkpoint widget is shown. As can be appreciated, one skilled in the art could develop a number of different widget formats in light of the present specification to graphically illustrate the progress and status of various checkpoints during a call. In this embodiment, a screen shot 1400 is shown that may be presented to an agent during a call.”), wherein the guidance data comprises a suggested new stage of the current workflow for the conversation (Neuer, Fig. 13B; Col. 31, lines 40-46: “The widget shown in FIG. 13A-13B illustrates that a number of checkpoints can be expected during the call. The widget indicates in real-time which checkpoints have or have not occurred as the call progresses. This provides an easy-to-comprehend status of the call, allowing an agent or supervisor to quickly view and understand how the call is progressing.”; Col. 34, lines 4-6: “The widgets illustrated in FIGS. 13A, 13B, and 14A illustrate checkpoints that are normally expected to occur on a call.”), wherein the guidance data is based, at least in part, on the transmission of the notification and the and the behavioral and lexical analysis on the additional audio data (Neuer, Col. 34, lines 19-32: “Turning to FIG. 14B, the same screen shot 1400 is shown but with an added icon and text 1470. This comprises a warning symbol and text that reflects the speech condition that was detected for this checkpoint. This allows feedback to the agent to reinforce speech that is not desired for whatever reason. In this embodiment, the icon and text 1470 remains displayed until the agent acknowledges the alert. Further, the icon and text 1470 are only displayed once the checkpoint has been detected. In other embodiments, the alert could be displayed for a predefined time period. In other embodiments, the information may be logged into a file recording the agent's checkpoints, or a notification message could be sent to the agent's supervisor or team leader.”).
McCord, Kumar, and Neuer are analogous arts because they each belong to the same field of endeavor of customer service systems. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the customer interaction and experience system of McCord, as modified by the user communication recording system of Kumar, to incorporate the teachings of Neuer to score an agent based on their interactions with a customer. The additional scoring system can be used to evaluate how well an agent is performing during customer interactions (Neuer, col. 27, lines 26-28). Evaluating the performance of employees allows a company and its employees to determine ways to improve their services. However, McCord, in view of Kumar and Neuer, fails to expressly recite wherein the one or more models are structured as a plurality of processing elements comprising a directed acyclic graph executed as a parallel algorithm in which each processing element of the plurality of processing elements performs an operation as a parallel process when a set of inputs becomes available to the one or more models; determine that the notification includes customer relationship management data; in response to the determination that the notification includes customer relationship management data, transmitting the notification to a customer relationship management device.
Griffin teaches determine that the notification includes customer relationship management data (Griffin, Fig. 2, ¶ [0032]: "It is then determined whether the query pertains to data available on a public cloud resource, or the query pertains to data available on a private cloud resource (step 212)"); in response to the determination that the notification includes customer relationship management data, transmitting the notification to a customer relationship management device (Griffin, Fig. 2, ¶ [0033]: " if it is determined that the query pertains to the data available on the private cloud resource, the query is interpreted by the using an enterprise specific model trained on at least one machine learning technique on data from the private cloud (step 216).").
McCord, Kumar, Neuer, and Griffin are analogous arts because they each belong to the same field of endeavor of data processing. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the customer interaction and experience system of McCord, as modified by the user communication recording system of Kumar and the agent scoring system of Neuer, to incorporate the teachings of Griffin to identify the type of data in a transfer, and send that data to an appropriate system. Sending CRM data, including user data, to a different system then general data allows for additional security and protection of sensitive data (Griffin, ¶ [0003]). It is important to protect a user’s sensitive data so that no one gains unauthorized access to sensitive information. However, McCord, in view of Kumar, Neuer, and Griffin, fails to expressly recite wherein the one or more models are structured as a plurality of processing elements comprising a directed acyclic graph executed as a parallel algorithm in which each processing element of the plurality of processing elements performs an operation as a parallel process when a set of inputs becomes available to the one or more models.
Sivasubramanian teaches wherein the one or more models are structured as a plurality of processing elements comprising a directed acyclic graph executed as a parallel algorithm in which each processing element of the plurality of processing elements performs an operation as a parallel process when a set of inputs becomes available to the one or more models (Sivasubramanian, Fig. 1, 114; [0068]: “Step functions workflow 114 may refer to a series of operations that are specified to run a job. The workflow may be specified in the job, either directly or indirectly. For example, the job may specify a set of capabilities of comprehend service 132 to utilize as part of the workflow. A step function workflow may include a series of steps, some of which may be executed in parallel and others which are to be executed sequentially. A workflow may specify a set of dependencies that describes how the workflow is to be executed. A workflow may be represented as a directed acyclic graph where nodes represent different steps and directed edges represent dependencies.”).
McCord, Kumar, Neuer, Griffin, and Sivasubramanian are analogous arts because they both belong to the same field of endeavor of customer service systems. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the customer interaction and experience system of McCord, as modified by the user communication recording system of Kumar, the agent scoring system of Neuer, and the data privacy system of Griffin, to incorporate the teachings of Sivasubramanian to incorporate models running in parallel to classify aspects of a conversation. This allows multiple processes to be done at one, whenever an input is ready (Sivasubramanian, [0068]), thus allowing for more efficient processing of data.
Claim(s) 20 is/are rejected under 35 U.S.C. 103 as being unpatentable over McCord, in view of Kumar, Neuer, Griffin, and Sivasubramanian, as applied to claim 19 above, and further in view of Ramachandran.
Regarding claim 20, the rejection of claim 19 is incorporated. McCord, in view of Kumar, Neuer, Griffin, and Sivasubramanian, discloses all of the elements of the current invention as stated above. However, McCord, in view of Kumar, Neuer, Griffin, and Sivasubramanian, fails to expressly recite wherein the guidance and scoring signals comprises guidance for interacting with a party to the call session.
Ramachandran teaches wherein the guidance and scoring signals comprises guidance for interacting with the second party during the conversation (Ramachandran, ¶ [0011]: "Embodiments of the present invention relate to a method and an apparatus for providing automated workflow guidance to an agent in a call center environment, for example, during a voice call between an agent and a customer of a business.").
McCord, Kumar, Neuer, Griffin, Sivasubramanian, and Ramachandran are analogous arts because they both belong to the same field of endeavor of customer service systems. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the customer interaction and experience system of McCord, as modified by the user communication recording system of Kumar, the agent scoring system of Neuer, the data privacy system of Griffin, and the customer contact analysis system of Sivasubramanian, to incorporate the teachings of Ramachandran to include models for detection context, call type, and call topic. These additional models allow for an automatic computer system to identify more aspects of a customer service call, which is important because "there exists a need for improved call center computing and management systems, which can provide real-time automated guidance on a workflow to be used to an agent” (Ramachandran, ¶ [0005]).
Conclusion
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to TYLER J BECKER whose telephone number is (703)756-1271. The examiner can normally be reached M-Th, 7:15am-5:45pm PT.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Daniel Washburn can be reached at (571) 272-5551. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/TYLER BECKER/ Examiner, Art Unit 2657
/DANIEL C WASHBURN/ Supervisory Patent Examiner, Art Unit 2657
