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 .
Continued Examination Under 37 CFR 1.114
A request for continued examination under 37 CFR 1.114, including the fee set forth in 37 CFR 1.17(e), was filed in this application after final rejection. Since this application is eligible for continued examination under 37 CFR 1.114, and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114. Applicant's submission filed on 1/12/26 has been entered.
Notice to Applicant
The following is a Non-Final Office action. In response to Examiner’s Final Rejection of 9/11/25, Applicant, on 1/12/26, amended claims. Claims 1-21 are pending in this application and have been rejected below.
Response to Amendment
Applicant’s amendments are acknowledged.
The previous 112b rejection is withdrawn in light of the amendment.
The 101 rejections are withdrawn in light of the amendments, as the claim is no longer directed to an abstract idea when viewing the claim in combination; and is also viewed as an improving computer technology (MPEP 2106.05a) and having “meaningful limitations” (MPEP 2106.05e). When viewing the claim in combination, the claim now requires a processor perform each step, where a first set of evaluation prompts are generated by an LLM, where chunks of interactions are evaluated by a machine learning model, and then a second set of training recommendation prompts are created to generate training recommendations using “machine learning” and querying a database using training categories from the 2nd set of training recommendations prompts.
Claim Rejections - 35 USC § 112
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 2, 5, 7, 9, 13, 16, and 18 are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
Claims 2, 13 recite the limitation " generating evaluation results comprises comparing one or more performance indicators identified in the interaction data items with threshold values for the performance indicators”; but claim 1 now recites “A method of evaluating performance in interactions… evaluating interaction data items of one or more interactions, wherein one or more of the evaluation prompts comprise threshold parameters.” There is insufficient antecedent basis for “threshold value” in claims 2, 13. It is unclear if this is a second set of thresholds, or the same as claim 1. Examiner is not sure what to suggest. It may be that claims 2, 13 are not further limiting the claim and need to be cancelled.
Claims 5, 16 recite the limitation " wherein generating evaluation results for the evaluation prompts comprises conducting a similarity search using the vector store”; but claim 1 now recites “computing a similarity between a vector representation of an evaluation prompt of the plurality of evaluation prompts and one or more vector representations of overlapping chunks of an interaction of the one or more interactions.” There is insufficient antecedent basis for “a similarity search” in claims 5, 16. It is unclear if this is a second similarity search, or if it is referring to the same similarity with vectors as in claim 1. Examiner is not sure what to suggest. It may be that claims 5, 16 are not further limiting the claim and need to be cancelled.
Claims 7, 18 also introduce “a similarity search.” They are rejected for the same reasons as claims 5, 16; as it appears the limitations are now in claim 1 possibly, or it’s referring to a second similarity search.
Claim 9 recites “wherein the evaluation prompts and training recommendation prompts are submitted to a large language machine learning model.” However, claim 1 is now amended to recite “the plurality of evaluation prompts are generated by a large language model (LLM) based on parameters extracted from an evaluation form.” It is unclear if claim 9 is referring to the same large language model or a different one. Examiner is not sure what to suggest. It may be that claim 9 is not further limiting the claim and need to be cancelled. Claim 9 is also unclear because claim 1 has evaluation prompts being generated “by a LLM”; in claim 9 though, the evaluation prompts and training recommendation prompts are “submitted to” a large language model.” It is unclear what is desired. There may be support for either situation, but it is unclear the way claim 1 and 9 currently recite the limitations what is happening.
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 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.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claims 1-21 are rejected under 35 U.S.C. 103 as being unpatentable over Arora (US 12,062,368) in view of Qazvinian (US 2025/0013963) and Hiray (US 2024/0395258)t.
Concerning claim 1, Arora discloses:
A method of evaluating performance in interactions and generating training recommendations based on the evaluated agent performance (Arora – see Col. 4, lines 54-67 - In at least one embodiment, contacts analytics service is integrated into a customer call service console and allows supervisors to conduct fast, full-text search on call and chat transcripts, discover themes and emerging trends from customer contacts, and improve agent performance with analytics-based coaching tools. See col. 5, lines 1-11 - Contacts analytics service may provide real-time analytics for both supervisors and agents during live calls which can provide actionable insights and suggestions to deliver improved customer support. Supervisors can use contacts analytics service's visual dashboard with call scoring to track all in-progress calls and intervene when customers are having a poor experience. Agents can use contacts analytics service's suggested answers to address live customer queries more effectively;
see also Qazvinian – see par 88 - e-learning systems can contain information about, for example, employee training courses, certifications, and/or skill development programs. By utilizing this data, the system can offer responses to prompts related to employee training history, skill profiles, or recommended learning opportunities), the method comprising, using one or more computer processors (Arora – see col. 73, lines 32-42 - includes a computer-readable storage medium (e.g., a hard disk, random access memory, read only memory, etc.) storing instructions that, if executed by a processor of the server, cause or otherwise allow the server to perform its intended functions):
creating, by one or more computer processors, a plurality of evaluation prompts for evaluating interaction data items of one or more interactions (Applicant’s [0117] as published and FIG. 6 state “ Evaluation prompts may be generated from evaluation prompt templates 602 and quality management evaluation forms 604 using machine learning… to implement LLM in applications may be used to generate evaluation prompts from evaluation prompt templates 602 and transcripts from interaction analytics 606C which may include interaction data items. Input to a LLM may be in form of a prompt, e.g. an evaluation prompt or a training recommendation prompt or may include input that may be derived from an interaction recording. For example, evaluation questions and evaluation parameters of evaluation forms such as “Number of elevated calls to a supervisor for an agent” may be extracted from an evaluation form and used in the generation of an evaluation prompt. The extraction of content present in evaluation forms may be analyzed using Gen AI-based LLMs provided by OpenAI and LangChain as a framework to implement LLM in applications
Arora – see col. 6, lines 57-67 - Contacts analytics service may be provide capabilities for agents and supervisors that are integrated into a contact center service's user experience (e.g., graphics-based consoles and interfaces). In at least some embodiments, supervisors have access to new and enhanced existing user interface elements within a contact center service that allow them to categorize conversations, setup call scoring, search historical contacts, derive themes, provide post-call agent coaching; see col. 9, lines 40-67 - For example, instead of supervisors listening to a randomly selected sample of calls and relying upon skewed customer satisfaction surveys, contacts analytics service can be used to analyze and categorize all calls. Supervisors can use contacts analytics service to review comments and/or feedback for specific portions of historic calls, categorize historic calls to determine compliance with different organizational rules or categories. See col. 42, lines 13-39, FIG. 8 - In at least one embodiment, a theme is discovered and then a pending new category is created in which a client can review a new category that was programmatically created through theme detection. For example, phrases 802 may be pre-filled to include one or more key phrases extracted from customer contacts that are highly correlated with a theme detected through semantic clustering; FIG. 8 shows the phrases searched for is “thanks for being a subscriber”; client may modify the pending new category; see col. 45, lines 6-20 - contacts analytics service may perform additional post-processing, such as assigning a sentiment score to portions of the audio contact and/or a sentiment score for the overall contact), wherein one or more of the evaluation prompts comprise threshold parameters, and wherein the plurality of evaluation prompts are generated … based on parameters extracted from an evaluation form (Arora see col. 49, lines 12-26 - As a second example, organizations (e.g., employees thereof) can adjust various settings to set threshold associated with sentiments and define actions or categories based on certain thresholds being exceeded. see col. 50, lines 34-53 - categories are displayed in the contact analysis dashboard, which in this illustrative example is depicted in the lower-right portion of FIG. 21. Themes may be categories that are manually generated by a rules that a client defines (e.g., in accordance with FIGS. 8-9). see col. 54, lines 25-55 - For example, a category to identify a potentially problematic call may rely on successive negative sentiment scores, loud volume, profanity uttered by the customer, utterances of the customer referencing competitor products/threats to cancel a subscription, and various combinations thereof. NLP service 2310 may be used to generate insights, which may include entity detection, sentiment analysis, and more, which are provided to contacts analytics service in any suitable format).
Qazvinian discloses:
creating, by one or more computer processors, a plurality of evaluation prompts for evaluating interaction data items of one or more interactions, wherein one or more of the evaluation prompts comprise threshold parameters, and wherein the plurality of evaluation prompts are generated “by a large language model (LLM)” based on parameters extracted from an evaluation form (Applicant’s specification [0109] as published states “ They may further include reasons for an evaluation result, may identify interaction data items that are above a threshold, e.g. meet or exceed expectations when compared to a threshold value for an interaction data item, may identify interaction data items that are below a threshold, e.g. don't meet expectations when compared to a threshold value for an interaction data item and can be summarized in a grade for an evaluation question within a certain scale, e.g. customer satisfaction for handling a call was rated 7 out of 10 (on a scale from 1 to 10, 10 being the highest and 1 being the lowest score).”
Qazvinian discloses entire limitation – see FIG. 1B – prompt submission, module 170; par 33 – machine learning module 185 may be a “generative AI system”; see par 37 - the received prompt is in natural language format, allowing users to express their prompts using everyday language and familiar language patterns (disclosing “extracted from an evaluation form”, which is input from a person). In some embodiments, to effectively process and understand the natural language input, the system leverages advanced natural language processing (hereinafter “NLP”) techniques. It may employ state-of-the-art AI models, such as, e.g., transformer-based models, which are trained on large datasets. see par 45 - In some embodiments, the system categorizes prompts into different sets based on their topics or intents, updates the categorized prompts based on user input to improve the accuracy and relevance of the categorization, and then allows users to use the categories as a guide to quickly generate prompts related to specific topics; see par 53-55, FIG. 2 –At step 220, the system generates an embedding representation of the received prompt; at least one of the generative AI models operates by processing the natural language prompt and transforming it into a dense numerical representation known as an embedding. This embedding captures the underlying meaning and nuances of the prompt, allowing for more accurate and context-aware responses. In some embodiments, the generative AI model may take the form of a large language model which has been trained on vast amounts of text data. see par 66 - once the candidate prompts are identified, the system compares their embeddings to the generated embedding representation of the current prompt. In some embodiments, various similarity metrics may be utilized, such as, for example, cosine similarity or Euclidean distance, to measure the similarity between embeddings. In some embodiments, the system may apply threshold-based filtering to exclude prompts with similarity scores below a certain threshold, ensuring that only highly similar prompts are considered. see par 102 - the generative AI system includes at least one general purpose large language model (hereinafter “LLM”)).
Arora and Qazvinian disclose:
generating, by one or more computer processors, evaluation results for the interaction data items using the plurality of evaluation prompts and machine learning (Applicant’s [0108] as published states “interaction data items can also be excerpts or snippets of interaction metadata and analytics results.” Applicant’s [0114] as published states “an interaction data item such as AHT may be a key performance indicator (KPI) that measures the average amount of time an agent spends on handling a customer interaction, e.g. including talk time, hold time, and after-call work.” [0132] as published states “] In the generation of evaluation results 738, one or more performance indicators identified in the interaction data items may be compared with pre-set threshold values for the performance indicators…focus areas can be parameters that can be derived from interaction data items, e.g. AHT (Average Handling Time).
Arora - Col. 3, lines 37-40 - Machine learning algorithms may be utilized to identify and tag turns in a conversation that mention an issue or call driver. Col. 3, lines 37-40 - Machine learning algorithms may be utilized to identify and tag turns in a conversation that mention an issue or call driver. see col. 49, lines 59-67, col. 50, lines 1-13 - A contact analysis dashboard 2100 surfaces various information at a glance to a supervisor. In various embodiments, contact analysis dashboard 2100 is a web-based UI. Contact analysis dashboard 2100 surfaces aggregate statistics at the top of the UI, in at least one embodiment, and displays one or more aggregate statistics. Average handle time (AHT) may refer to the average length of a customer interaction, which includes hold time, talk time, and after-call work (ACW). ), wherein the generating of the evaluation results comprises computing a similarity between a … representation of an evaluation prompt of the plurality of evaluation prompts and one or more … representations of overlapping chunks of an interaction of the one or more interactions, wherein the … representation of the evaluation prompt and the one or more … representations of the overlapping chunks of the interaction are created by a machine learning model (Arora - Col. 5, lines 51-67 ; Col. 7, lines 16-31 - contacts analytics service uses highly accurate speech transcription technology to transcribe calls and automatically indexes call transcripts and chat-based interactions so that they are searchable in the contact center service console. Col. 6, lines 10-40 - Supervisors can track agent compliance with defined categorization rules that provide parameters for how agents interact with customers; Supervisors can also track agent performance by defining categories that organize customer contacts based on content and characteristics such as silence duration, sentiment, talk speed, and interruptions. See col. 18, lines 7-49 - Sentiment analysis 120A may refer to analyzing text (e.g., a turn, being a portion of a text-based transcript of an audio recording) and determining one or more characteristics of the call. For example, sentiment analysis 120A of a statement may generate a sentiment score that indicates a sentiment of the statement in question was positive, neural, negative, or mixed. A sentiment score may be generated based on successive sentiments of a speaker—for example, if a customer's sentiment of a first turn is positive, it may be assigned an initial sentiment score value of +1; entity detection 120B may be used on a call transcript to identify products, dates, events, locations, organizations (e.g., competitors), persons, quantities, titles, and more. see col. 43, lines 55-67 - FIG. 11 illustrates a contact search result page 1100. In at least one embodiment, contact search result page 1100 is provided to a client of a service provider in response to a client search request with a specified set of search parameters. In at least one embodiment, search parameters may include various parameters such as … one or more keywords or phrases. Searches may be performed using NLP techniques so that literal as well as semantic matches are returned), wherein the generating of the evaluation results comprises comparing the interaction data items with the threshold parameters (Applicant’s [0140] as published states “ An evaluation using machine learning such as Gen AI in combination with a large language model and word embedding may allow evaluating an interaction. For example, each question of an evaluation form may be embedded into an evaluation prompt that is used to query an interaction transcript with the help of machine learning, e.g. Gen AI. A generated evaluation result may include items for a question such as, for example: an answer to a question, a reason for the answer, positive feedback, negative feedback.”
Arora discloses the limitations based on broadest reasonable interpretation in light of the specification - see col. 52, lines 56-67, col. 53, lines 1-11 - A categorization may be applied to the angry customer illustrated in FIG. 22 based on a negative-trending customer sentiment, which may be based on a successive run of negative sentiments and/or a downward trend of customer sentiment from positive to negative. Such a category may be presented to supervisor 2214 via customer contact service 2212, which may surface a notification or a dashboard may have a dedicated widget or UI element for surfacing potentially problematic calls).
Arora discloses that supervisors can track agent compliance with defined categorization rules with sentiment and talk speed (See col. 6) where sentiment analysis on a turn/portion of a transcript occurs (See col. 18, lines 7-49); searching for set of parameters including keywords/phrases (See co. 43); and using a dashboard to retrieve “problematic”/negative sentiment interactions (See col. 52). Arora then further discloses identifying “turns” of historical contact center records that are labeled with identified issues and/or call drivers (See col. 23, lines 59-67).
Qazvinian disclose:
generating, by one or more computer processors, evaluation results for the interaction data items using the plurality of evaluation prompts and machine learning, wherein the generating of the evaluation results comprises computing a similarity between a “vector representation” of an evaluation prompt of the plurality of evaluation prompts and one or more “vector representations of overlapping chunks of an interaction of the one or more interactions, wherein the vector representation of the evaluation prompt and the one or more vector representations of the overlapping chunks of the interaction” are created by a machine learning model (Qazvinian – See par 49 - the system leverages machine learning algorithms to identify relevant factors and variables that contribute to the prompt's topic. It analyzes, e.g., historical data, performance metrics, employee feedback, and other relevant information to generate recommendations tailored to the organization's specific needs. see par 50 - generated recommendations can cover topics, such as, e.g., talent development initiatives, performance improvement methods; see par 62-64 – step 230 in FIG. 2 is similarity search using embedding representation of prompt; see par 66 - once the candidate prompts are identified, the system compares their embeddings to the generated embedding representation of the current prompt. In some embodiments, various similarity metrics may be utilized. see par 70 - Each prompt in the database can be represented as a vector in a high-dimensional space, capturing its inherent characteristics. During the similarity search process, the system compares the generated embedding representation of the received prompt with the pre-existing embedding representations in the vector database. see par 71 - In some embodiments, the system retrieves a set of similar prompts by selecting the top-k nearest neighbors based on the similarity scores. These similar prompts can serve as a resource for generating responses to the received prompt or providing insights based on historical data. see par 88 – HRIS data source may store performance evaluations. , the system can offer responses to prompts related to employee training history, skill profiles, or recommended learning opportunities).
Arora and Qazvinian disclose:
creating, by one or more processors, training recommendation prompts including the evaluation results, wherein one or more of the training recommendation prompts comprise training categories ([084 For example, training recommendation prompts may be created for evaluation results of an agent may be identified for evaluation results, that have not met a certain threshold for a specific interaction data item. E.g. in case that an agent’s average handling time for interactions was evaluated as exceeding 10 minutes, a training recommendation prompt may be created that requests the provision of training recommendations that may enable an agent to improve on an average handling time for customer agent interactions.
Arora - See col. 6, lines 1-22 - In at least some embodiments, contacts analytics service includes agent coaching capabilities that enables supervisors to find opportunities to increase their agents' effectiveness—for example, contacts analytics service may generate a graphical illustration for past calls that makes it easy for supervisors to spot issues. Supervisors can track agent compliance with defined categorization rules that provide parameters for how agents interact with customers; See col. 10, lines 36-47 - contacts analytics service may transcribe call audio in real-time and detect instances where an agent says “I don't know” or “I don't handle that” to detect instances in which an agent's responses may cause customers frustration. In at least some embodiments, customer/agent tone (e.g., customer yelling at agent) or failure by agent to adhere to script and compliance procedures may be flagged to a supervisor dashboard to provide supervisors more transparency into how customer issues are being resolved; Col. 29, lines 60-67, col. 30, lines 1-10 - NLP service 610 may perform entity and phrase detection to identify important aspects of customer contacts. NLP insights may be encoded in an output file or response that is provided to contacts analytics service 606. In some cases, NLP service 610 parses contacts data and generates suggestions to questions or issues presented by customers as part of a real-time agent assistance feature. For example, NLP service 610 may parse a customer's turn to detect phrases and entities that indicate that the customer is having trouble with a product and is requesting a return. NLP service 610 may generate suggested responses, such as troubleshooting steps, which may be surfaced to an agent via customer contact service 604.
Qazvinian See par 45 - the system categorizes prompts into different sets based on their topics or intents, updates the categorized prompts based on user input to improve the accuracy and relevance of the categorization, and then allows users to use the categories as a guide to quickly generate prompts related to specific topics. see par 52 - prompts may be example prompts, such as, e.g., follow-up prompts or suggested recommendations that the user may wish to submit in this session but hasn't submitted yet, that the generative AI system can suggest to the user. The user may optionally select one of these prompts to be submitted as the next prompt to be submitted. Thus, in some embodiments, one or more suggested prompts can be displayed for the user during the generative AI session; see par 82 - data warehouse employs a robust prompt processing engine capable of efficiently executing the executable expression. In some embodiments, this engine may utilize prompt optimization techniques, such as, for example, prompt rewriting, indexing, and caching, to expedite the retrieval and processing of relevant data. For example, in response to a prompt regarding employee attrition rates, the system can formulate an optimized prompt that leverages precomputed statistics or indexes within the data warehouse to quickly obtain the required information; see par 114 - the system utilizes the conversation context to generate follow-up prompts. Based on the previous interactions and the current state of the conversation, the system can identify areas where further clarification or elaboration is required. In some embodiments, to generate appropriate response outputs to the follow-up prompts, the system can utilize the same generative AI model employed for processing the initial prompt. By leveraging the contextual information from the conversation history, the model can generate responses that build upon previous interactions.); and
generating, by one or more of the processors, training recommendations from the training categories using the training recommendation prompts and machine learning, wherein the generating of the training recommendations comprises querying a database using the training categories (Arora –Col. 9, lines 40-67 - Accordingly, an organization can use contacts analytics service to more effectively train supervisors and/or agents. Supervisors can use contacts analytics service to review comments and/or feedback for specific portions of historic calls, categorize historic calls to determine compliance with different organizational rules or categories; Col. 10, lines 48-59 - contacts analytics service provides real-time agent assistance. Contacts analytics service may use artificial intelligence and machine learning to provide in-call assistance to agents based on real-time call audio which is being transcribed and analyzed to generate suggestions to agents to help them better solve customers' issues. In at least some embodiments, real-time transcripts of calls and/or chats are provided to Kendra which can then provide specific answers or give a list of relevant documents from the company's knowledge base (e.g., using a document ranking feature) to help an agent more quickly locate an answer to the customer's specific question. see Col. 57, lines 13 – 42, FIG. 25 – transcribe audio data (2510); execute NLP to generate metadata (2512 – disclosing evaluation results), determine whether categories match metadata (2514 – disclosing training recommendation prompts), generate suggestion based on transcript, metadata, and categories (2516 – disclosing last portion of limitation)).
Arora discloses “ In at least some embodiments, contacts analytics service includes agent coaching capabilities that enables supervisors to find opportunities to increase their agents' effectiveness—for example, contacts analytics service may generate a graphical illustration for past calls that makes it easy for supervisors to spot issues and share feedback with agents by commenting on specific portions of the conversation” (See col. 6, lines 1-20) and taking sentiment scores for customer contacts to use machine learning models to identify pain points and give “additional information presented to agents in real-time” (See col. 49, lines 13-26).
Qazvinian and Hiray also disclose:
generating training recommendations from “training categories” using the training recommendation prompts and machine learning wherein the generating of the training recommendations comprises querying a database using the “training categories” (Qazvinian See par 88 - one or more e-learning systems can be leveraged as data sources within the data warehouse. Such e-learning systems can contain information about, for example, employee training courses, certifications, and/or skill development programs. By utilizing this data, the system can offer responses to prompts related to employee training history, skill profiles, or recommended learning opportunities
To any extent Qazvinian is viewed as just giving general training recommendations, Hiray discloses the limitations as well:
Hiray – see par 26 - , the chatbots correspond to generative AI tools that generate assistive content to support representatives. the chatbots monitor active conversations between representatives and third-parties, analyze the dialog and the context associated with the dialog, generate customized content based on the analyzed dialog and context, and present… , and/or alert the representatives about behaviors that deviate from best practices or that may improve the outcome of the conversation. see par 116 - automated conference monitoring system 100 uses AI/ML techniques to evaluate the aggregate contextual reports, and generate actionable data based on the evaluation; actionable data may provide… deriving best practices for improving the overall effectiveness of the team; see par 129 - conference monitoring system 100 uses one or more AI/ML techniques to analyze the recorded set of conversations, determine common contextual trends or patterns that exist with a certain frequency amongst the recorded set of conversations with a common rating, score, or classification, and generate (at 1104) a coaching model based on the common contextual trends or patterns, defined criteria, and/or manager-selected contextual trackers 603. In this manner, automated conference monitoring system 100 generates coaching models that are customized according to preferences of each manager. The coaching model may include a set of contextual trackers 603 for common context associated with best practices or desired behavioral paradigms or common context associated with undesired practices or behavioral paradigms).
Arora, Qazvinian, and Hiray are analogous art as they are directed to improving employees (see Arora Abstract, Col. 4 - agents interacting/conversing with customers; insights from customer conversations; Qazvinian Abstract, par 87-88; Hiray Abstract - agents interacting/conversing with customers). 1) Arora disclose having text search (col. 6, 8) and searches for phrases and then natural language processing to look for phrases with similar syntactic meanings (See col. 42-43) and generates suggestions to solve customer issues and for agents as part of real-time agent assistance (See col. 10; col. 29-30) where various suggestions can be based made on transcript, metadata, and categories (FIG. 25, 2510-2516; col. 57). Qazvinian improves upon Arora by having examples of best practices or desired behavior based on classified contextual trackers, using a LLM (large language model) for prompting the model with specific topics based on initial information (See par 37) and then processes the prompt (see par 53-55); representing similarity based on vectors/embeddings for generating responses (See par 70-71), and then creating follow-up prompts and prompt re-writing to obtain required information (See par 52, 82, 114); and provide e-learning, training courses, skill development programs related to skill profiles or learning opportunities (See par 88). One of ordinary skill in the art would be motivated to further include using LLMs, creating follow-up prompts or prompt re-writing to obtain required information, and e-learning and skill development to efficiently improve upon the searches for semantic matches using NLP in Arora. Hiray improves upon Arora and Qazvinian by disclosing using generative AI to give information to improve outcome of a conversation (See par 26, 116) and generating coaching models customized according to preference of managers where contact is associated with best practices (par 129). One of ordinary skill in the art would be motivated to further include using context for examples of different best practices to efficiently improve upon the searches for semantic matches using NLP in Arora.
Accordingly, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the natural language processing discovering opportunities for agent coaching in Arora, to further have LLM with prompts and prompt re-writing that is used for retrieving information and used for learning opportunities as disclosed in Qazvinian, and to further have LLM/generative AI used for coaching recommended best practices for different context/classifications as disclosed in Hiray, since the claimed invention is merely a combination of old elements, and in combination each element merely would have performed the same function as it did separately, and one of ordinary skill in the art would have recognized that the results of the combination were predictable and there is a reasonable expectation of success.
Concerning independent claim 12, Arora, Qazvinian, and Hiray disclose:
A system for evaluating agent performance in interactions and generating training recommendations for agents based on the evaluated agent performance, the system comprising:
a computing device (Arora – see col. 73, lines 32-42 - processor);
a memory (Arora – see col. 73, lines 32-42 - a computer-readable storage medium (e.g., a hard disk, random access memory, read only memory, etc.) storing instructions); and
a processor, the processor configured to (Arora – see col. 73, lines 32-42 - includes a computer-readable storage medium (e.g., a hard disk, random access memory, read only memory, etc.) storing instructions that, if executed by a processor of the server, cause or otherwise allow the server to perform its intended functions);
The remaining limitations are similar to claim 1 above. It would be obvious to combine Arora, Qazvinian, and Hiray for the same reasons as claim 1.
Concerning independent claim 20, Arora, Qazvinian, and Hiray disclose:
A method of rating agent performance in interactions and providing suggestions for improvement (Arora [as in claim 1]– see Col. 4, lines 54-67; col. 5, lines 1-11; see also Qazvinian – see par 88), the method comprising:
creating one or more prompts for the evaluation of interaction data items in one or more interactions (Arora discloses the limitations based on broadest reasonable interpretation in light of the specification – [same as claim 1] - see col. 6, lines 57-67; See col. 9, lines 40-67; col. 42, lines 13-39, FIG. 8; col. 45, lines 6-20);
generating performance results for the interaction data items using the one or more prompts for the evaluation of interaction data items and machine learning, wherein… (Arora -same as claim 1] - Col. 3, lines 37-40; see col. 49, lines 59-67, col. 50, lines 1-13 - Contact analysis dashboard 2100 surfaces aggregate statistics at the top of the UI, in at least one embodiment, and displays one or more aggregate statistics. Average handle time (AHT);
Qazvinian – See par 49-50, 62-64, 66, 70-71);
creating coaching recommendation prompts for the performance results (similar to claim 1 above - Arora - See col. 6, lines 1-22 - at least some embodiments, contacts analytics service includes agent coaching capabilities that enables supervisors to find opportunities to increase their agents' effectiveness…; col. 10, lines 36-47;
see also Qazvinian See par 45; see par 52; see par 82; see par 114); and
generating coaching recommendations from coaching categories using the coaching recommendation prompts and machine learning… ([similar to claim 1 above - Arora – see col. 8, lines 23-49; Col. 9, lines 40-67; Col. 10, lines 48-59; Col. 40, lines 52-67; col. 42, lines 13-39
see also Qazvinian See par 88; Hiray par 26, 116, 129).
The remaining limitations are similar to claim 1 above. It would be obvious to combine Arora and Hiray for the same reasons as claim 1.
Concerning claims 2 and 13, Arora, Qazvinian, and Hiray disclose:
A method according to claim 1, wherein generating evaluation results comprises comparing one or more performance indicators identified in the interaction data items with threshold values for the performance indicators (Arora [as in cl. 1] - Col. 49, lines 12-26 - As a second example, organizations (e.g., employees thereof) can adjust various settings to set threshold associated with sentiments and define actions or categories based on certain thresholds being exceeded. For example, a run of N negative sentiments coupled with an overall negative sentiment score may be categorized as a bad interaction and a supervisor may be notified (e.g., after the fact or in real-time); see col. 54, lines 25-55 - For example, a category to identify a potentially problematic call may rely on successive negative sentiment scores, loud volume, profanity uttered by the customer, utterances of the customer referencing competitor products/threats to cancel a subscription, and various combinations thereof. NLP service 2310 may be used to generate insights, which may include entity detection, sentiment analysis
See also Hiray – see par 129 - Automated conference monitoring system 100 uses one or more AI/ML techniques to analyze the recorded set of conversations, determine common contextual trends or patterns that exist with a certain frequency amongst the recorded set of conversations with a common rating, score, or classification, and generate (at 1104) a coaching model based on the common contextual trends or patterns, defined criteria, and/or manager-selected contextual trackers 603; see par 147 - Process 1300 includes selecting (at 1304) a particular member of the particular group to evaluate. The selection (at 1304) may be made periodically (e.g., quarterly review, annual review, etc.), based on specific events associated with the particular member (e.g., the particular member having a conversion rate that is below a specified threshold)).
It would be obvious to combine Arora and Qazvinian and Hiray for the same reasons as claim 1.
Concerning claims 3 and 14, Arora, Qazvinian, and Hiray disclose:
A method according to claim 1, wherein the training recommendation prompts are created for the evaluation results that comprise one or more performance indicators that are lower than thresholds for the one or more performance indicators (Arora [as in cl. 1] - Col. 49, lines 12-26 - As a second example, organizations (e.g., employees thereof) can adjust various settings to set threshold associated with sentiments and define actions or categories based on certain thresholds being exceeded. For example, a run of N negative sentiments coupled with an overall negative sentiment score may be categorized as a bad interaction and a supervisor may be notified (e.g., after the fact or in real-time); see col. 54, lines 25-55 - For example, a category to identify a potentially problematic call may rely on successive negative sentiment scores, loud volume, profanity uttered by the customer, utterances of the customer referencing competitor products/threats to cancel a subscription, and various combinations thereof. NLP service 2310 may be used to generate insights, which may include entity detection, sentiment analysis
Hiray – see par 129 - Automated conference monitoring system 100 uses one or more AI/ML techniques to analyze the recorded set of conversations, determine common contextual trends or patterns that exist with a certain frequency amongst the recorded set of conversations with a common rating, score, or classification, and generate (at 1104) a coaching model based on the common contextual trends or patterns, defined criteria, and/or manager-selected contextual trackers 603; see par 147 - Process 1300 includes selecting (at 1304) a particular member of the particular group to evaluate. The selection (at 1304) may be made periodically (e.g., quarterly review, annual review, etc.), based on specific events associated with the particular member (e.g., the particular member having a conversion rate that is below a specified threshold).
It would be obvious to combine Arora, Qazvinian, and Hiray for the same reasons as claim 1.
Concerning claims 4 and 15, Arora, Qazvinian, and Hiray disclose:
A method according to claim 1, wherein creating evaluation prompts comprises creating a vector store for the interaction data items (Applicant’s specification [0091] as filed states “In a similarity search, a service may identify chunks of an interaction recording that are most relevant to an evaluation prompt created by a user by identifying a similarity of chunks, e.g. chunks that may include interaction data items, and a prompt, e.g. an evaluation prompt. This may be done by computing a similarity between a vector representation of a prompt and vectors of chunks, e.g. to identify a semantic similarity between chunks of an interaction recording and the evaluation prompt.”
Arora discloses the limitations based on broadest reasonable interpretation in light of the specification - col. 4, lines 1-16 – key phrases clustered to detect common issues that are shared; clustering may involve representing key phrases as N-dimensional points which are then clustered based on relative distance to each other, where points that are closer to each other have more similar semantic meaning than points that are farther from each other. Clusters may be identified using a deep learning based learning algorithm; see col. 25, lines 4-22 - Graph 400 illustrates a two-dimensional visualization of key phrases that are clustered based on similar contents. While a two-dimensional visualization is shown in FIG. 4, higher dimensions are utilized to encode semantic meaning of different key phrases across N-dimensional spaces. See col. 29, lines 48-67 - Transcripts, chat logs, and other text-based contacts data may be provided by contacts analytics service 606 to NLP service 610 (See FIG. 27); NLP service performs various natural language processing techniques such as those in FIG. 1. For example, chat logs may be organized into turns and each turn may be provided to NLP service 610 to determine a sentiment of the turn. Sentiments can be used to determine the overall mood and progression of a conversation—for example, if the sentiment of a customer starts out as negative and trends positive after successive turns, then that contact may be considered a good contact. However if a customer's sentiment trends negative and ends negative at the end of a customer contact, that may indicate that there was a difficulty with the contact and may require additional investigation by a supervisor. NLP service 610 may perform entity and phrase detection to identify important aspects of customer contacts. NLP insights may be encoded in an output file or response that is provided to contacts analytics service 606. For example, NLP service 610 may parse a customer's turn to detect phrases and entities that indicate that the customer is having trouble with a product and is requesting a return.
Qazvinian see par 62-64 – step 230 in FIG. 2 is similarity search using embedding representation of prompt; see par 66 - once the candidate prompts are identified, the system compares their embeddings to the generated embedding representation of the current prompt. In some embodiments, various similarity metrics may be utilized; see par 70 - Each prompt in the database can be represented as a vector in a high-dimensional space, capturing its inherent characteristics. During the similarity search process, the system compares the generated embedding representation of the received prompt with the pre-existing embedding representations in the vector database)
It would be obvious to combine Arora, Qazvinian, and Hiray for the same reasons as claim 1.
Concerning claims 5 and 16, Arora, Qazvinian, and Hiray disclose:
A method according to claim 4, wherein generating evaluation results for the evaluation prompts comprises conducting a similarity search using the vector store (Arora – col.. 20, lines 25-40 - As a third stage, clustering is performed on some or all key phrases that were extracted from a set of customer contacts (e.g., collected over a period of time that is either pre-defined or specified by a customer) and semantically clustered into themes. In at least some embodiments, a deep learning based learning algorithm is utilized by theme detection data plane 136 to perform clustering. In some embodiments, an unsupervised learning algorithm is used to identify clusters which are surfaced as themes. In at least one embodiment, a clustering algorithm computes N-dimensional coordinates to a set of key phrases based on semantic meaning and clusters are assigned to non-overlapping subsets of the set of key phrases. Col. 25, lines 4-57 - FIG. 4 illustrates an example with three clusters 402-406 of themes on graph 400 that are programmatically identified through clustering, according to at least one embodiment. In some embodiments, each key phrase of a set of contacts data is represented as a point on a N-dimensional plane (e.g., two-dimensional Cartesian plane). The distance between two points on such a plane represents how semantically similar the points are to each other; In at least some embodiments, clustering analyses used to identify clusters may be based on one or more of nearest neighbor search, … or other types of cluster analysis techniques to determine clusters based on key phrase data.
Qazvinian – See par 49 - the system leverages machine learning algorithms to identify relevant factors and variables that contribute to the prompt's topic. It analyzes, e.g., historical data, performance metrics, employee feedback, and other relevant information to generate recommendations tailored to the organization's specific needs. see par 50 - generated recommendations can cover topics, such as, e.g., talent development initiatives, performance improvement methods; see par 69 - the system performs the similarity search by comparing the generated embedding representation of the received prompt with pre-existing embedding representations of previously submitted prompts stored in a vector database. This can enable the system to identify similar prompts that have been submitted before and leverage the corresponding responses to those similar prompts to provide accurate and relevant responses for currently received prompts. see par 70 - Each prompt in the database can be represented as a vector in a high-dimensional space, capturing its inherent characteristics. During the similarity search process, the system compares the generated embedding representation of the received prompt with the pre-existing embedding representations in the vector database. see par 71 - In some embodiments, the system retrieves a set of similar prompts by selecting the top-k nearest neighbors based on the similarity scores. These similar prompts can serve as a resource for generating responses to the received prompt or providing insights based on historical data).
It would be obvious to combine Arora, Qazvinian, and Hiray for the same reasons as claim 1 and 4.
Concerning claims 6 and 17, Arora, Qazvinian, and Hiray disclose:
A method according to claim 1, wherein the training categories are separated into training data items that are stored in a vector store ([0106] as filed - Coaching categories may be focus areas and behaviors. Focus areas can be parameters that can be derived from interaction data items, e.g. AHT. A coaching category may, thus include for example improving the AHT of an agent. Behaviors, e.g. active listening, effective questioning may be behavior characteristics of an agent that are linked to a focus area and improvement in the behavior characteristics of an agent during an interaction may improve a focus area. For example, to improve an AHT of an agent, an agent may receive a training that improves active listening and effective questioning of an agent Arora – see Col. 6, lines 10-40 -Supervisors can also track agent performance by defining categories that organize customer contacts based on content and characteristics such as silence duration, sentiment, talk speed, and interruptions. In at least some embodiments, contacts analytics service provides real-time assistance to supervisors and/or agents. In at least some embodiments, real-time supervisor assistance allows a supervisor to monitor call center analytics data in real-time, which may be aggregated across an entire call center, to specific product or service lines, or even a view onto a specific agent. Col. 29, lines 47-67, Col. 30, lines 1-10 - NLP service 610 may perform entity and phrase detection to identify important aspects of customer contacts. NLP insights may be encoded in an output file or response that is provided to contacts analytics service 606. In some cases, NLP service 610 parses contacts data and generates suggestions to questions or issues presented by customers as part of a real-time agent assistance feature. For example, NLP service 610 may parse a customer's turn to detect phrases and entities that indicate that the customer is having trouble with a product and is requesting a return. NLP service 610 may generate suggested responses, such as troubleshooting steps, which may be surfaced to an agent via customer contact service 604.
Qazvinian see par 70-71 - various similarity metrics can be employed, such as, e.g., cosine similarity or Euclidean distance, to measure the similarity between the vectors. see par 71 - In some embodiments, the system retrieves a set of similar prompts by selecting the top-k nearest neighbors based on the similarity scores. These similar prompts can serve as a resource for generating responses to the received prompt or providing insights based on historical data. See par 88 - one or more e-learning systems can be leveraged as data sources within the data warehouse. Such e-learning systems can contain information about, for example, employee training courses, certifications, and/or skill development programs. By utilizing this data, the system can offer responses to prompts related to employee training history, skill profiles, or recommended learning opportunities
See also Hiray – See par 73-74 – semantic similarity on discussed topics for assigning topic to context tracker for segments or snippets; See par 105 – conference monitoring system 100 stores context for conferences in a database, repository, or other storage, and retrieves based on conference and/or context based on identifiers; See par 146 - Process 1300 includes generating (at 1302) a coaching model based on common contextual trends or patterns identified within monitored conversations of a particular group that produced a desired outcome or that exemplify best practices and/or desired behavioral paradigms for that particular group. See par 148-149 – FIG. 13 shows ‘retrieve’ contextual reports; query transcripts associated with monitored conversations; - Process 1300 includes comparing (at 1308) the context from contextual reports 600 to the common contextual trends or patterns of the coaching model, and selecting (at 1310) a first set of contextual trackers 603 for a first set of context from contextual reports 600 that correspond to the best practices and/or the desired behavioral paradigms in the coaching model. See par 152 - Controller 109 may track the particular member's actions or behaviors over time and after each performance evaluation to determine if the particular member has implemented the requested changes or if additional coaching or training is needed. The instructions may include identification of specific topics to discuss or behavioral changes. For instance, controller 109 may notify the particular member when they are speaking too fast, when they are speaking for too long of a duration, when questions are not posed to the customer, and/or when they interrupt the customer).
It would be obvious to combine Arora, Qazvinian, and Hiray for the same reasons as claim 1 and 4.
Concerning claims 7 and 18, Arora, Qazvinian, and Hiray disclose:
A method according to claim 6, wherein generating training recommendations comprises conducting a similarity search using the training data items vector store (Arora – similar to claim 5 - col. 20, lines 25-40; Col. 25, lines 4-57;
Qazvinian see par 69 - the system performs the similarity search by comparing the generated embedding representation of the received prompt with pre-existing embedding representations of previously submitted prompts stored in a vector database. This can enable the system to identify similar prompts that have been submitted before and leverage the corresponding responses to those similar prompts to provide accurate and relevant responses for currently received prompts. see par 70 - Each prompt in the database can be represented as a vector in a high-dimensional space, capturing its inherent characteristics. During the similarity search process, the system compares the generated embedding representation of the received prompt with the pre-existing embedding representations in the vector database. see par 71 - In some embodiments, the system retrieves a set of similar prompts by selecting the top-k nearest neighbors based on the similarity scores. These similar prompts can serve as a resource for generating responses to the received prompt or providing insights based on historical data. see par 88 – HRIS data source may store performance evaluations. , the system can offer responses to prompts related to employee training history, skill profiles, or recommended learning opportunities
See also Hiray – See par 73-74 – semantic similarity on discussed topics for assigning topic to context tracker for segments or snippets; See par 105 [as in claim 6], See par 146 - Process 1300 includes generating (at 1302) a coaching model based on common contextual trends or patterns … See par 148-149 – FIG. 13 shows ‘retrieve’ contextual reports; query transcripts associated with monitored conversations;… See par 152 - Controller 109 may track the particular member's actions or behaviors over time and after each performance evaluation to determine if the particular member has implemented the requested changes or if additional coaching or training is needed. The instructions may include identification of specific topics to discuss or behavioral changes. For instance, controller 109 may notify the particular member when they are speaking too fast, when they are speaking for too long of a duration, when questions are not posed to the customer, and/or when they interrupt the customer).
It would be obvious to combine Arora, Qazvinian, and Hiray for the same reasons as claim 1 and 4 and 6.
Concerning claims 8 and 19, Arora, Qazvinian, and Hiray disclose:
A method according to claim 1, wherein the machine learning comprises generative artificial intelligence (Qazvinian – see par 33 - , the machine learning module 185 may constitute at least a portion of a generative AI system used in various embodiments. see par 55 - t least one of the generative AI models operates by processing the natural language prompt and transforming it into a dense numerical representation known as an embedding.
Hiray – see par 26 - chatbots correspond to generative AI tools that generate assistive content to support representatives. In some such embodiments, the chatbots monitor active conversations between representatives and third-parties, analyze the dialog and the context associated with the dialog, generate customized content based on the analyzed dialog and context, … and/or alert the representatives about behaviors that deviate from best practices or that may improve the outcome of the conversation; See par 127 - Automated conference monitoring system 100 may also generate coaching models that are customized for each team member. The coaching models provide objective and corroborating evidence for the weaknesses or performance deficiencies of individual team members during performance reviews or coaching sessions; see par 128-129 – automated conference monitoring system 100 generates coaching models customized according to preferences of each manager).
It would be obvious to combine Arora, Qazvinian, and Hiray for the same reasons as claim 1.
Concerning claim 9, Arora, Qazvinian, and Hiray disclose:
A method according to claim 8, wherein the evaluation prompts and training recommendation prompts are submitted to a large language machine learning model (Qazvinian see par 102 - the generative AI system includes at least one general purpose large language model (hereinafter “LLM); The LLM generates a response output and determines a type of response that would be ideal for the nature of the prompt; For example, one specialized language model may be fine-tuned specifically for responding to recruitment-related prompts, another may be fine-tuned specifically for responding to employee attrition prompts, another may be fine-tuned specifically for responding to diversity and inclusion prompts, and so forth. see par 103 - if the response is evaluated to be incorrect, then the system sends the prompt to the LLM, the LLM generates a response
Hiray – see par 100 – automated conference monitoring system 100 trains and generates language learning models (LLMs) …context-aware speech-to-text LLM may be created and used to transcribe conferences involving customer support representatives. Each LLM is trained to accurately detect and transcribe the custom phrasing, jargon, sentence structure, and/or other conversational nuances that the different departments may use in discussions with other conference participants).
It would be obvious to combine Arora, Qazvinian, and Hiray for the same reasons as claim 1, 4, and 6.
Concerning claim 10, Arora, Qazvinian, and Hiray disclose:
A method according to claim 1, wherein the training recommendations are based on previously generated evaluation results for the agent (Arora – see col. 6, lines 1-30 - In at least some embodiments, contacts analytics service includes agent coaching capabilities that enables supervisors to find opportunities to increase their agents' effectiveness. Supervisors can track agent compliance with defined categorization rules that provide parameters for how agents interact with customers—for example, a supervisor may review call transcripts to determine how often an agent greets the customer in a call, which may be part of an agent handbook that guides agent behavior to provide a more pleasant and uniform customer experience. Supervisors can also track agent performance by defining categories that organize customer contacts based on content and characteristics such as silence duration, sentiment, talk speed, and interruptions; see col. 9, lines 60-67, Col. 10, lines 1-10 - Supervisors can use contacts analytics service to review comments and/or feedback for specific portions of historic calls, categorize historic calls to determine compliance with different organizational rules or categories. Agents can receive objective feedback provided by supervisors in at least some embodiments. Supervisors can, in at least some embodiments, mark specific calls with a thumbs up or thumbs down and/or comments and an agent can listen to the portion of the call where the supervisor provided feedback for taking more concrete corrective measures. In some embodiments, contacts analytics service provides an interface which supervisors can use to assign labels/tags for recurring searches (e.g., mapping to topics like customer churn and agent script adherence).
It would be obvious to combine Arora and Hiray for the same reasons as claim 1.
Concerning claim 11, Arora, Qazvinian, and Hiray disclose:
A method according to claim 1, wherein each evaluation prompt of the plurality of evaluation prompts comprises interaction data items which are present in previously created evaluation prompts (Arora - see col. 6, lines 57-67 - Contacts analytics service… In at least some embodiments, supervisors have access to new and enhanced existing user interface elements within a contact center service that allow them to categorize conversations, setup call scoring, search historical contacts, derive themes, provide post-call agent coaching, and various suitable combination thereof. See col. 8, lines 23-49 - fast full-text search that can be used by supervisors to diagnose problems such as customer churn by searching for past conversations where customers have expressed frustration with the company's products or mentioning cancelling their services; Organizations may use this capability to investigate the magnitude of known issues by searching through transcripts of past customer conversations and categorizing the calls to identify common issues. see col. 43, lines 16-39 - In some embodiments, contact search page 1000 includes additional search parameters not illustrated in FIG. 10, such as capabilities to search by categories. Search results may be displayed in accordance with some or all of FIGS. 11-14; FIG. 10 shows a search for keywords of “account is locked,” “can’t access my account”; see col. 43, lines 40-67 - Searches may be performed using NLP techniques so that literal as well as semantic matches are returned; Contact search result page 1100 may display a set of common themes that are detected for the search parameters. Common themes may refer to keywords and phrases that are positively correlated with search parameters.)
Concerning claim 21, Arora, Qazvinian, and Hiray disclose:
The method of claim 1, wherein the interaction data items comprise excerpts of a voice recording or a video recording (Arora – col. 13, lines 52-67 - Client data store 106 may refer to an electronic data store that an organization uses to store contact data. Contact data may refer to audio recordings of calls between agents and customers, chat logs of online conversations between agents and customers, video interactions between agents and customers, and more. see col. 17, lines 5-25 - transcription of text-based contacts data may be skipped, as the data is already in a text-based format. However, in at least one embodiment, audio-based contacts data (e.g., video and audio recordings) may be transcribed using a speech-to-text service 132. In at least one embodiment, speech-to-text service 132 receives audio data (e.g., in the form of an audio or video file) and generates a text-based transcript of the audio data), and wherein the threshold parameters comprise a handling time of interactions (see col. 49, lines 59-67, col. 50, lines 1-13 - A contact analysis dashboard 2100 surfaces various information at a glance to a supervisor. In various embodiments, contact analysis dashboard 2100 is a web-based UI. Contact analysis dashboard 2100 surfaces aggregate statistics at the top of the UI, in at least one embodiment, and displays one or more aggregate statistics. Average handle time (AHT) may refer to the average length of a customer interaction, which includes hold time, talk time, and after-call work (ACW)).
Response to Arguments
Applicant's arguments filed 1/21/26 have been fully considered but they are not persuasive and/or are moot in view of the new rejections.
With regards to 103, Applicant’s arguments are moot in view of new rejections necessitated by the amendments.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to IVAN R GOLDBERG whose telephone number is (571)270-7949. The examiner can normally be reached 830AM - 430PM.
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, Anita Coupe can be reached at 571-270-3614. 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.
/IVAN R GOLDBERG/Primary Examiner, Art Unit 3619
Prosecution Insights