DETAILED ACTION
Notice of 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 Arguments
Applicant’s amendments and arguments filed 03/16/2026, with respect to claim(s) 1-20 have been fully considered. Applicant amended claims 1, 3, 9, 16-20, cancelled claims 4, 7, 8, 10 and added new claims 21-24.
Applicant’s arguments in pages 11-17, filed 03/16/2026, with respect to 35 U.S.C 101 rejections of Claims 1-20 have been fully considered but they are not persuasive. Applicant argued that the additional features in the amended claims describe operations that involve computational processing of textual data through a large language model (e.g., at least one trained neural network comprising billions of parameters) that applies complex matrix operations, attention mechanisms, and transformer architectures to process natural language input. It is neither practical nor feasible for a human to perform the above computational operations entirely in the human mind or with pen and paper. Rather, the operations can only be reasonably completed by a computer and are specifically tailored to a computing environment. Examiner respectfully disagrees. The amended claimed limitations describe multiple steps of determination of customer service resolution. It describe generating issue resolution indicators based on interaction transcript, purpose, one or more issues and resolution prompt. Confidence scores have been calculated based on the probability for both resolved and unresolved status. Narratives or summaries for both cases ( resolved and unresolved) are displayed/presented. The method described above could be done by a human with pen and paper. Throughout the independent claims, first and second large language model are mentioned as “ using” them. No unique or specialized features or training to achieve those features of the large language models to perform the above methods are described in the claims. “Large language model” are mentioned as an additional element, as performing generic computer functions that are well-understood, routine and conventional activities amount to no more than implementing the abstract idea with a computerized system.
Applicant further argued, that the claims are patent-eligible because the claims integrate any alleged abstract idea into a practical application, as set forth in the 2019 revised Patent Subject Matter Eligibility Guidance. Examiner respectfully disagrees. Using the first and second large language models as additional elements, as generic computing element, did not reflect an improvement in the functioning of a computer, or an improvement to other technology or technical field. Applicant mentioned few paragraph of the specification which shows some features of large language models which brings improvement in the practical application. But none of the features are reflected in the claim language. Claim language is showing large language model as additional element, as performing generic computer functions. The use of a computer does not preclude performance of the invention via pen and paper or in a person’s mind. Also, the use of a computer or other machinery in its ordinary capacity to perform a task or simply adding a general purpose computer to an abstract idea, does not integrate a judicial exception into a practical application. Here the computer is the machine that is merely an object on which the method operates, which does not integrate the exception into a practical application or provide significantly more.
Applicant again argued that the amended claims are similar to disputed claims in Ex parte Desjardins and analogous claims are eligible for patenting. Examiner respectfully disagrees. In Ex parte Desjardins, the claims recite the steps of training a machine learning model by elaborating on first training data, second training data, how the values of plurality of parameters are adjusted to optimize performance of machine learning model. The claim language recites the specialized features on how the training of the model is done and how the improvement is achieved. In contrast, the claims of current application recite the steps of determining customer service resolution by using large language model. The claim language is silent about what type of specialized features or how the large language models are trained to perform the task to make improvement to the technology. The claims are not similarly constitute as the claims in Exparte Desjardins.
Applicant further argued that the mere disclosure of a few known examples of large language models in the field is not an admission or explicit statement that demonstrates the well-understood, routine, or conventional nature of large language models. To support a conclusion that an additional element (or combination of additional elements) is well- understood, routine, or conventional activity, the MPEP requires that the examiner must make a factual determination that includes at least one of “[a] citation to an express statement in the specification or to a statement made by an applicant during prosecution that demonstrates the well- understood, routine, conventional nature of the additional element(s)". Examiner would like to mention that para.[0030] in specification says “ the issue resolution indication process implements one or more large language models (LLMs) to analyze the interaction transcript 102 and generate content for the issue resolution indication. Examples of LLMs include, but are not limited to, OpenAI’s ChatGPT, NeMO™ LLM from NVIDIA®, LLaMa from Meta®, BERT from Google®, CLAUDE™ from Anthropic A.I., and FLAN-T5 form Google®. The process described herein can implement one or more LLMs currently developed or that may be developed in the future”, which shows the LLMs used to process the steps in claims are generic LLM. Applicant again argued that even if large language models could appropriately be characterized as generic computer components, arguendo, the MPEP teaches that generic computer components "can still integrate or amount to significantly more when considered in combination with the other elements of the claim. But the claim language didn’t illustrate any improvement in the technology by using generic LLM with other generic computer components, which is not sufficient to amount to significantly more than the judicial exception. Thus, the 35 U.S.C 101 rejections of Claims 1, 2, 5, 6, 9, 11-20 have been maintained. 35 U.S.C 101 rejection of claim 3 has been withdrawn in view of the amended claim filed on 03/16/2026. Newly added claims 21, 22 are method and apparatus claims performing the similar steps in claim 3 and therefor, there is no 35 U.S.C. 101 rejection for those claims.
Applicant’s arguments and amendments filed 03/16/2026, with respect to claim(s) 1-3, 5, 6, 9, 11-20 under 35 U.S.C. 103 have been fully considered but are not persuasive. Applicant argued that the amended claimed limitation recite additional features, including those recited in previously presented claims 3 and 4 are not taught by any of the previous references. Examiner respectfully disagrees. The amended claimed limitation has some features from previous claim 3, 4 and crossed out features from claim 9, which were previously rejected by reference Ghoche. Examiner believes Ghoche does teach the purpose is generated using a second large language model and summarizes a set of one or more intents detected in the interaction transcript;
Compact Prosecution
In order to advance prosecution of this case, the examiner reached out to the attorney with a proposed examiner’s amendment to add the amended claim 3 with the independent claims or at least some of the technical features from claim 3 to independent claims to overcome 35 U.S.C. 101, but no agreement was reached to place the case in condition for allowance. Please note this is an attempt to provide suggestions to further advance prosecution but has not been searched. An updated search would be required if agreed to.
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1, 2, 5, 6, 9, 11-20, 23 and 24 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Independent claims 1, 16 recite “obtaining an interaction transcript from the interaction between a first entity and a second entity”; “obtaining a resolution prompt”; “determining one or more issues presented to the first entity by the second entity”; “a first large language model based on the interaction transcript, the one or more issues, a purpose corresponding to the interaction transcript, and the resolution prompt, one or more issue resolution indications wherein each issue resolution indication corresponds and wherein the purpose is generated using a second large language model and summarizes a set of one or more intents detected in the interaction transcript”; “generating, for each of the one or more issue resolution indications indicating resolved status, a first narrative summarizing one or more actions implemented to resolve the one or more issues”; “generating, for each of the one or more issue resolution indications indicating unresolved status, a second narrative summarizing a reason the one or more issues are unresolved”; “generating, with the first large language model for the one or more issue resolution indications indicating resolved status, a first confidence score corresponding to one of: a probability that the resolved status is in fact indicating an issue, corresponding to the issue resolution indication indicating resolved status, is resolved”, “or a probability that the unresolved status is in fact indicating an issue, corresponding to the issue resolution indication indicating unresolved status, is unresolved “; “and outputting the first narrative or the second narrative with each respective one of the one or more issue resolution indications”. The limitations above as drafted, is a process that, under its broadest reasonable interpretation, covers a mental process, as this could be performed in the human mind or with the aid of pen and paper.
The limitation of " obtaining ... ", “determining..”, “invoking…”, “generating…”, “ outputting…” as drafted covers mental activities. More specifically, a human can obtain a transcript of an interaction between a customer and a customer service representative, determine one or more issues from that transcript and a resolution, can generate an indication of each issue resolution, either resolved or unresolved, the issue resolution indication can be calculated as a probability and a confidence score can be calculated based on the probability, where the score can indicate the issue resolution status, can generate summary of how the issue was resolved or unresolved , what steps were taken and present or keep record of both summaries corresponding to the issues. All the steps above are examples of observation and evaluation that could be performed in the human mind or with the aid of pencil and paper.
The claims recite the additional limitation of “large language model”, claim 16 recite “processor”, “memory” for performing the method. All those are recited at a high level of generality and are recited as performing generic computer functions routinely used in computer applications. The current specification in paragraph [0030] specifies large language model can be OpenAI’s ChatGPT, NeMO™ LLM from NVIDIA®, LLaMa from Meta®, BERT from Google®, CLAUDE™ from Anthropic A.I., and FLAN-T5 form Google®., which is not sufficient to amount to significantly more than the judicial exception. Specification in paragraphs [0098],[0134], clearly specifies “processor”, “memory” as performing generic computer functions that are well-understood, routine and conventional activities amount to no more than implementing the abstract idea with a computerized system. The claims as drafted, are not patent eligible.
Thus, taken alone, the additional elements do not amount to significantly more than the above identified judicial exception (the abstract idea). Looking at the limitations as an ordered combination adds
nothing that is not already present when looking at the elements taken individually. There is no indication
that the combination of elements improves the functioning of a computer or improves any other technology. Their collective functions merely provide conventional computer implementation. Claims 1
and 16 are therefore not drawn to eligible subject matter as they are directed to an abstract idea without
significantly more than the abstract idea.
Independent claim 18 recites “obtaining an interaction transcript from the interaction between a first entity and a second entity”; “obtaining a resolution prompt”; “invoking a first large language model with a first input comprising at least the interaction transcript”; “determining, with the first large language model, one or more issues presented to the first entity by the second entity”; “generating, with the first large language model based on the first input comprising the interaction transcript, the one or more issues, a purpose corresponding to the interaction transcript,” “and the resolution prompt, one or more issue resolution indications, wherein each issue resolution indication corresponds wherein the purpose is generated using a second lame language model and summarizes a set of one or more intents detected in the interaction transcript, and wherein the one or more issue resolution indications comprises at least one of a resolution status or a resolution narrative”; “and outputting the one or more issue resolution indications”. The limitations above as drafted, is a process that, under its broadest reasonable interpretation, covers a mental process, as this could be performed in the human mind or with the aid of pen and paper.
The limitation of " obtaining ... ", “determining..”, “invoking…”, “generating…”, “ outputting…” as drafted covers mental activities. More specifically, a human can obtain a transcript of an interaction between a customer and a customer service representative, determine one or more issues from that transcript and a resolution, can generate an indication of each issue resolution, either resolved or unresolved, generate summary and present the issue resolution indicator corresponding to the issues. All the steps above are examples of observation and evaluation that could be performed in the human mind or with the aid of pencil and paper.
The claim recites the additional limitation of “large language model”, for performing the method, which is recited at a high level of generality and are recited as performing generic computer functions routinely used in computer applications. The current specification in paragraph [0030] specifies large language model can be OpenAI’s ChatGPT, NeMO™ LLM from NVIDIA®, LLaMa from Meta®, BERT from Google®, CLAUDE™ from Anthropic A.I., and FLAN-T5 form Google®., which is not sufficient to amount to significantly more than the judicial exception. Generic computer functions that are well-understood, routine and conventional activities amount to no more than implementing the abstract idea with a computerized system. The claim as drafted, is not patent eligible.
Thus, taken alone, the additional element does not amount to significantly more than the above identified judicial exception (the abstract idea). Looking at the limitations as an ordered combination adds
nothing that is not already present when looking at the elements taken individually. There is no indication
that the combination of elements improves the functioning of a computer or improves any other technology. Their collective functions merely provide conventional computer implementation. Claim 18 is therefore not drawn to eligible subject matter as they are directed to an abstract idea without
significantly more than the abstract idea.
Claims 2 and 17 recite the additional limitation of “wherein determining the one or more issues presented to the first entity by the second entity comprises determining the one or more issues with the first large language model based on a first input comprising at least the interaction transcript and a prompt instructing the first large language model to identify the one or more issues” , where determining the issues from a transcript of conversation, could be performed in the human mind or with the aid of pen and paper. The claims recite additional limitation of large language model, which is specified in specification in paragraph [0030] as OpenAI’s ChatGPT, NeMO™ LLM from NVIDIA®, LLaMa from Meta®, BERT from Google®, CLAUDE™ from Anthropic A.I., and FLAN-T5 form Google®. which is generic and can be any neural network or any machine learning model, which is not sufficient to amount to significantly more than the judicial exception. The claims 2 and 17 as drafted, are not patent eligible.
Claim 5 recites “wherein determining the one or more issues presented to the first entity by the second entity comprises determining the one or more issues with a third large language model based on a first input comprising at least the interaction transcript”, where determining the issue from the interaction transcript is an evaluation, observation and could be performed in the human mind or with the aid of pen and paper. The claim recites additional limitation of large language model, which is specified in specification in paragraph [0030] as OpenAI’s ChatGPT, NeMO™ LLM from NVIDIA®, LLaMa from Meta®, BERT from Google®, CLAUDE™ from Anthropic A.I., and FLAN-T5 form Google®. which is generic and can be any neural network or any machine learning model, which is not sufficient to amount to significantly more than the judicial exception. The claim 5 as drafted, is not patent eligible.
Claim 6 recites “wherein: generating the first narrative comprises invoking the first large language model to generate the first narrative, and generating the second narrative comprises invoking the first large language model to generate the second narrative”, to generate the summary could be performed in the human mind or with the aid of pen and paper. The claim recites additional limitation of large language model, which is specified in specification in paragraph [0030] as OpenAI’s ChatGPT, NeMO™ LLM from NVIDIA®, LLaMa from Meta®, BERT from Google®, CLAUDE™ from Anthropic A.I., and FLAN-T5 form Google®. which is generic and can be any neural network or any machine learning model, which is not sufficient to amount to significantly more than the judicial exception. The claim 6 as drafted, is not patent eligible.
Claim 9 recites “further comprising:
Claim 11 recites “further comprising: receiving a plurality of first narratives corresponding to a plurality of interactions; converting textual data structure of the plurality of first narratives into numerical vectors embeddings; discerning, with a categorization component processing the numerical vectors embeddings, one or more categories that are present within the plurality of first narratives; and labeling, with a label generation component, the one or more categories with a keyphrase”, to determine a summary of interaction, converting them in vector, put them in different category and labeling with keyphrase could be done with the aid of pen and paper. The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception, as claim 11 does not recite any additional limitations. The claim as drafted, is not patent eligible.
Claim 12 recites “wherein the keyphrase generated by the label generation component is a phrase extracted from an overlapping portion of the plurality of first narratives categorized within each of the one or more categories”, to determine that the keyphrase generated from the overlapped portion of multiple summary, is an evaluation, observation and could be performed in the human mind or with the aid of pen and paper. The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception, as claim 12 does not recite any additional limitations. The claim as drafted, is not patent eligible.
Claim 13 recites “wherein the keyphrase generated by the label generation component is an abstraction based on the plurality of first narratives categorized within each of the one or more categories”, to determine that the keyphrase generated from the summary is an abstraction, is an evaluation, observation and could be performed in the human mind or with the aid of pen and paper. The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception, as claim 13 does not recite any additional limitations. The claim as drafted, is not patent eligible.
Claim 14 recites “wherein obtaining the interaction transcript comprises: receiving an audio recording of the interaction between the first entity and the second entity; and generating, with a fourth large language model configured for speech recognition processing, the interaction transcript”, where the interaction transcription can be obtained from a recoded conversation and could be performed in the with the aid of pen and paper. The claim recites additional limitation of large language model, which is specified in specification in paragraph [0030] as OpenAI’s ChatGPT, NeMO™ LLM from NVIDIA®, LLaMa from Meta®, BERT from Google®, CLAUDE™ from Anthropic A.I., and FLAN-T5 form Google®. which is generic and can be any neural network or any machine learning model, which is not sufficient to amount to significantly more than the judicial exception. The claim 14 as drafted, is not patent eligible.
Claim 15 recites “wherein the one or more issue resolution indications comprises a Boolean status”, where resolution indication can be yes or no, which is an evaluation, observation and could be performed in the human mind or with the aid of pen and paper. The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception, as claim 15 does not recite any additional limitations. The claim as drafted, is not patent eligible.
Claim 19 recites “further comprising: generating, with the first first large language model, for each of the one or more issue resolution indications indicating unresolved status, a second narrative summarizing a reason the one or more issues are unresolved; and outputting the first narrative or the second narrative corresponding to each respective one of the one or more issue resolution indications”, where generating a summary for issues with resolved status, generating a summary for issues with unresolved status and outputting or presenting those reports, could be performed with the aid of pen and paper. The claim recites additional limitation of large language model, which is specified in specification in paragraph [0030] as OpenAI’s ChatGPT, NeMO™ LLM from NVIDIA®, LLaMa from Meta®, BERT from Google®, CLAUDE™ from Anthropic A.I., and FLAN-T5 form Google®. which is generic and can be any neural network or any machine learning model, which is not sufficient to amount to significantly more than the judicial exception. The claim 19 as drafted, is not patent eligible.
Claim 20 recites “further comprising: generating, with the first large language model for the one or more issue resolution indications indicating resolved status, a first confidence score corresponding to one of: a probability that the resolved status is in fact indicating an issue, corresponding to the issue resolution indication indicating resolved status, is resolved, or a probability that the unresolved status is in fact indicating an issue, corresponding to the issue resolution indication indicating unresolved status, is unresolved”; “and determining whether the first confidence score is greater than or equal to a threshold “; “storing the issue resolution indication in one or more memories based on the determination that the first confidence score is greater than or equal to the threshold”; “and generating, with a second first large language model based on the interaction transcript and the resolution prompt, one or more additional issue resolution indications and a second confidence score, based on the determination that the first confidence score is not greater than or equal to the threshold”, where generating an issue resolution indications of both resolved and unresolved status, such as confidence score and based on a threshold of the confidence score, saving the issue resolution indications, generating addition issue resolution indications, scores and comparing with threshold, could be performed with the aid of pen and paper. The claim recites additional limitation of large language model, which is specified in specification in paragraph [0030] as OpenAI’s ChatGPT, NeMO™ LLM from NVIDIA®, LLaMa from Meta®, BERT from Google®, CLAUDE™ from Anthropic A.I., and FLAN-T5 form Google®. which is generic and can be any neural network or any machine learning model, which is not sufficient to amount to significantly more than the judicial exception. The claim 20 as drafted, is not patent eligible.
Claim 23 recites “wherein the second large language model is different from the first large language model”, where mentioning that the first and second large language models are different , is an observation, evaluation, could be performed in human mind or with the aid of pen and paper. The claim recites additional limitation of large language model, which is specified in specification in paragraph [0030] as OpenAI’s ChatGPT, NeMO™ LLM from NVIDIA®, LLaMa from Meta®, BERT from Google®, CLAUDE™ from Anthropic A.I., and FLAN-T5 form Google®. which is generic and can be any neural network or any machine learning model, which is not sufficient to amount to significantly more than the judicial exception. The claim 23 as drafted, is not patent eligible.
Claim 24 recites “wherein the first large language model comprises one of a small language model (SLM) or a low complexity large language model”, where describing what the first large language model comprised of, is an observation, evaluation, could be performed in human mind or with the aid of pen and paper. The claim recites additional limitation of large language model, which is specified in specification in paragraph [0030] as OpenAI’s ChatGPT, NeMO™ LLM from NVIDIA®, LLaMa from Meta®, BERT from Google®, CLAUDE™ from Anthropic A.I., and FLAN-T5 form Google®. which is generic and can be any neural network or any machine learning model, which is not sufficient to amount to significantly more than the judicial exception. The claim 24 as drafted, is not patent eligible.
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 of this title, 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.
Claims 1, 2, 5, 6, 9 and 15-20 are rejected under 35 U.S.C. 103 as being unpatentable over Li et al. ( US 20250005589 A1), hereinafter referenced as Li, in view of Ghoche et al. (US 20240386213 A1), hereinafter referenced as Ghoche.
Regarding Claim 1, Li teaches a method for providing an issue resolution indication for an interaction, comprising: obtaining an interaction transcript from the interaction between a first entity and a second entity ( Li: Para.[0111],[ 0126]-[0128], Fig. 5, incident data gathering tool 502 collect incident data ( interaction transcript) which include querying for and receiving the incident data. The collected incident data can include any data, electronically obtained from any system, tool, or software and can be in textual form. The obtained incident information may include chat messages exchanged between responders and others regarding the incident);
obtaining a resolution prompt ( Li: Para.[0129], Fig.5, The classification tool 504 prompts the classifier to perform a binary classification task, such as with respect to a reportable criterion C and a set of incident data D essentially states, 'does any of D describe/is about/include the reportable criterion C?');
generating, with a first large language model based on the interaction transcript, the one or more issues, a purpose corresponding to the interaction transcript, and the resolution prompt, one or more issue resolution indications, wherein each issue resolution indication corresponds ( Li: Para.[0117], Fig. 4, the status reporting software 416 can be used to generate incident status report ( issue resolution indication) by using incident data ( includes the interaction and issues), training data, and prompts as inputs to a language model, such as a generative artificial intelligence model or a large language model, to obtain incident status reports);
generating, for each of the one or more issue resolution indications indicating resolved status, a first narrative summarizing one or more actions implemented to resolve the one or more issues ( Li: Para.[0130], [0132], Fig. 5, Table 1, summarization tool 506 causes the summarization model 514 to extract a summary from the incident information where the summary is exclusively focused on the identified reportable criteria. As such, the summarization tool 506 can obtain the summary from the incident status report ( issue resolution indication) using only the subset of the incident information that correspond to the reportable criteria confirmed to be present in the collected incident information. Para. [0022], the incident can be a resolved incident);
generating, for each of the one or more issue resolution indications indicating unresolved status, a second narrative summarizing a reason the one or more issues are unresolved ( Li: Para.[0130], [0132], Fig. 5, Table 1, summarization tool 506 causes the summarization model 514 to extract a summary from the incident information where the summary is exclusively focused on the identified reportable criteria. As such, the summarization tool 506 can obtain the summary from the incident status report ( issue resolution indication) using only the subset of the incident information that correspond to the reportable criteria confirmed to be present in the collected incident information. Para. [0022], the incident can be an unresolved incident);
and outputting the first narrative or the second narrative with each respective one of the one or more issue resolution indications ( Li: Para.[0152], Fig. 10, at 1010 an incident status report ( summary) received from the language model is transmitted to one or more users (e.g. ,stakeholders).
Li while teaching the method of claim 1, fails to explicitly teach the claimed, determining one or more issues presented to the first entity by the second entity; and wherein the purpose is generated using a second large language model and summarizes a set of one or more intents detected in the interaction transcript; generating, with the first large language model for the one or more issue resolution indications indicating resolved status, a first confidence score corresponding to one of: a probability that the resolved status is in fact indicating an issue, corresponding to the issue resolution indication indicating resolved status, is resolved, or a probability that the unresolved status is in fact indicating an issue, corresponding to the issue resolution indication indicating unresolved status, is unresolved;
However, Ghoche does teach the claimed, determining one or more issues presented to the first entity by the second entity ( Ghoche: Para.[0069]-[0072], Fig. 1,determining customer issues with the AI augmented customer support module 140, where issues are presented at customer support application by the customer ( second entity) to the agents ( first entity));
and wherein the purpose is generated using a second large language model and summarizes a set of one or more intents detected in the interaction transcript ( Ghoche: Para.[0223], Fig. 47, an intent/topic is detected from the customer ticket, including an analysis of customer texts in block 4700. If a customer's ticket/text message indicated that their intent is to cancel an order, the workflow for “cancel order” is triggered, and autonomous AI chatbot 4704 and large language model 4706 interact to implement the cancel order workflow. Para.[0216], Fig. 42, by processing the text in an abstractive summarizer, the summary “unable to run payroll” may be mapped to the closest topic in a granular taxonomy, such as a topic for “payroll problem.”).
generating, with the first large language model for the one or more issue resolution indications indicating resolved status, a first confidence score corresponding to one of: a probability that the resolved status is in fact indicating an issue, corresponding to the issue resolution indication indicating resolved status, is resolved, [or a probability that the unresolved status is in fact indicating an issue, corresponding to the issue resolution indication indicating unresolved status, is unresolved] ( Ghoche: Para.[0087]-[0089], a confidence can be generated for a particular answer ( resolved status) which was generated by a particular macro rather than from other macros. BERT or other transformer based ML techniques can be used);
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate Ghoche’s teaching of an autonomous customer support Chatbot Agent utilizing a large language model to aid in implementing a workflow to solve a customer issue, into the system and method of smart incident status updates, taught by Li, because, this would improve overall system performance by using an autonomous AI chatbot that interacts with a large language model to implement an improved workflow.( Ghoche, Para.[0199],[0221]).
Claim 16 is an apparatus claim, comprising: one or more memories comprising processor-executable instructions; and one or more processors configured to execute the processor-executable instructions ( Li: Para.[0079], [0080], Fig. 2, memory 204, processor 202, may further include one or more data storage 210, which may further include program code, data, algorithms, and the like, for use by a processor, such as the processor 202 to execute and perform actions), performing the steps in method claim 1 above and as such, claim 16 is similar in scope and content to claim 1 and therefore, claim 16 is rejected under similar rationale as presented against claim 1 above.
Regarding Claim 2, Li in view of Ghoche teach the method of claim 1. Ghoche further teaches, wherein determining the one or more issues presented to the first entity by the second entity comprises, determining the one or more issues with the first large language model based on a first input comprising at least the interaction transcript and a prompt instructing the first large language model to identify the one or more issues ( Ghoche: Para.[0233], Fig. 49, in 4912 large language model is prompted, which may include information on the conversation. In 4914, a determination is made of actions and response for the autonomous AI chatbot).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate Ghoche’s teaching of an autonomous customer support Chatbot Agent utilizing a large language model to aid in implementing a workflow to solve a customer issue, into the system and method of smart incident status updates, taught by Li, because, this would improve overall system performance by using an autonomous AI chatbot that interacts with a large language model to implement an improved workflow.( Ghoche, Para.[0199],[0221]).
Claim 17 is an apparatus claim performing the steps in method claim 2 above and as such, claim 17 is similar in scope and content to claim 2 and therefore, claim 17 is rejected under similar rationale as presented against claim 2 above.
Regarding Claim 5, Li in view of Ghoche teach the method of claim 1. Ghoche further teaches, wherein determining the one or more issues presented to the first entity by the second entity comprises determining the one or more issues with a third large language model based on a first input comprising at least the interaction transcript ( Ghoche: Para.[0233], Fig. 49, in 4912 large language model is prompted, which may include information on the conversation. In 4914, a determination is made of actions and response for the autonomous AI chatbot).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate Ghoche’s teaching of an autonomous customer support Chatbot Agent utilizing a large language model to aid in implementing a workflow to solve a customer issue, into the system and method of smart incident status updates, taught by Li, because, this would improve overall system performance by using an autonomous AI chatbot that interacts with a large language model to implement an improved workflow.( Ghoche, Para.[0199],[0221]).
Regarding Claim 6, Li in view of Ghoche teach the method of claim 1. Li further teaches, wherein: generating the first narrative comprises invoking the first large language model to generate the first narrative ( Li: Para.[0117],[0132], Fig. 5, The summarization tool 506 transmits a summarization request to a language model, such a summarization model 514, to obtain the status report. The summarization request includes instructions to the large language model to use the identified reportable criteria. The summarization tool 506 causes the summarization model 514 to extract a summary ( first narrative) from the incident information where the summary is exclusively focused on the identified reportable criteria),
and generating the second narrative comprises invoking the first large language model to generate the second narrative ( Li: Para.[0117],[0132], Fig. 5, The summarization tool 506 transmits a summarization request to a language model, such a summarization model 514, to obtain the status report. The summarization request includes instructions to the large language model to use the identified reportable criteria. The summarization tool 506 causes the summarization model 514 to extract a summary ( second narrative) from the incident information where the summary is exclusively focused on the identified reportable criteria).
Regarding Claim 9, Li in view of Ghoche teach the method of claim 1. Ghoche further teaches, further comprising:
determining whether the first confidence score is greater than or equal to a threshold ( Ghoche: Para.[0087], a confidence ( first) can be generated for a particular answer ( resolved status) which was generated by a particular macro rather than from other macros. Para.[0088], BERT or other transformer based ML techniques can be used. Para.[0124], a template answer exceeds a selected threshold);
storing the issue resolution indication in one or more memories based on the determination that the first confidence score is greater than or equal to the threshold ( Ghoche: Para.[0266], storing in memory),
and generating, with a second first large language model based on the interaction transcript and the resolution prompt, one or more additional issue resolution to the threshold ( Ghoche: Para.[0114],generating confidence ( second) for partial completing a response answer. Para.[0124], the template answer doesn’t exceed threshold, the question of the ticket is routed to a human agent to resolve).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate Ghoche’s teaching of an autonomous customer support Chatbot Agent utilizing a large language model to aid in implementing a workflow to solve a customer issue, into the system and method of smart incident status updates, taught by Li, because, this would improve overall system performance by using an autonomous AI chatbot that interacts with a large language model to implement an improved workflow.( Ghoche, Para.[0199],[0221]).
Regarding Claim 15, Li in view of Ghoche teach the method of claim 1. Li further teaches, wherein the one or more issue resolution indications comprises a Boolean status ( Li: Para.[0029], incident status updates ( issue resolution indication) amounts to asking the language model simple, binary questions regarding a corpus of data ( e.g., incident data) and then crafting a specific summarization request based on the answers therewith focusing the language model on the most pertinent portions of the corpus of data).
Regarding Claim 18, Li teaches a method for providing an issue resolution indication for an interaction, comprising: obtaining an interaction transcript from the interaction between a first entity and a second entity ( Li: Para.[0111],[ 0126]-[0128], Fig. 5, incident data gathering tool 502 collect incident data ( interaction transcript) which include querying for and receiving the incident data. The collected incident data can include any data, electronically obtained from any system, tool, or software and can be in textual form. The obtained incident information may include chat messages exchanged between responders and others regarding the incident);
obtaining a resolution prompt ( Li: Para.[0129], Fig.5, The classification tool 504 prompts the classifier to perform a binary classification task, such as with respect to a reportable criterion C and a set of incident data D essentially states, 'does any of D describe/is about/include the reportable criterion C?');
invoking a first large language model with a first input comprising at least the interaction transcript ( Li: Para.[0117],[0129], using the incident data ( interaction transcript) as input to the large language model);
generating, with the first large language model based on the first input comprising the interaction transcript, the one or more issues, a purpose corresponding to the interaction transcript and the resolution prompt, one or more issue resolution indications, wherein each issue resolution indication corresponds wherein the purpose is generated using a second large language model and summarizes a set of one or more intents detected in the interaction transcript,] and wherein the one or more issue resolution indications comprises at least one of a resolution status or a resolution narrative ( Li: Para.[0117], Fig. 4, the status reporting software 416 can be used to generate incident status report ( issue resolution indication) by using incident data, training data, and prompts as inputs to a language model, such as a generative artificial intelligence model or a large language model, to obtain incident status reports. Para.[0130], [0132], Fig. 5, Table 1, summarization tool 506 causes the summarization model 514 to extract a summary from the incident information where the summary is exclusively focused on the identified reportable criteria. As such, the summarization tool 506 can obtain the summary from the incident status report ( issue resolution indication) using only the subset of the incident information that correspond to the reportable criteria confirmed to be present in the collected incident information.);
and outputting the one or more issue resolution indications ( Li: Para.[0152], Fig. 10, at 1010 an incident status report ( issue resolution indication) received from the language model is transmitted to one or more users (e.g. ,stakeholders)).
Li while teaching the method of claim 1, fails to explicitly teach the claimed, determining, with the first large language model, one or more issues presented to the first entity by the second entity; wherein the purpose is generated using a second large language model and summarizes a set of one or more intents detected in the interaction transcript.
However, Ghoche does teach the claimed, determining one or more issues presented to the first entity by the second entity ( Ghoche: Para.[0069]-[0072], Fig. 1,determining customer issues with the AI augmented customer support module 140, where issues are presented at customer support application by the customer ( second entity) to the agents ( first entity). Para.[0246], AI model can be a large language model);
wherein the purpose is generated using a second large language model and summarizes a set of one or more intents detected in the interaction transcript ( Ghoche: Para.[0223], Fig. 47, an intent/topic is detected from the customer ticket, including an analysis of customer texts in block 4700. If a customer's ticket/text message indicated that their intent is to cancel an order, the workflow for “cancel order” is triggered, and autonomous AI chatbot 4704 and large language model 4706 interact to implement the cancel order workflow. Para.[0216], Fig. 42, by processing the text in an abstractive summarizer, the summary “unable to run payroll” may be mapped to the closest topic in a granular taxonomy, such as a topic for “payroll problem.”).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate Ghoche’s teaching of an autonomous customer support Chatbot Agent utilizing a large language model to aid in implementing a workflow to solve a customer issue, into the system and method of smart incident status updates, taught by Li, because, this would improve overall system performance by using an autonomous AI chatbot that interacts with a large language model to implement an improved workflow.( Ghoche, Para.[0199],[0221]).
Regarding Claim 19, Li in view of Ghoche teach the method of claim 18. Li further teaches, further comprising:
generating, with the first ( Li: Para.[0117],[0132], Fig. 5, The summarization tool 506 transmits a summarization request to a language model, such a summarization model 514, to obtain the status report. The summarization request includes instructions to the large language model to use the identified reportable criteria. The summarization tool 506 causes the summarization model 514 to extract a summary ( first narrative) from the incident information where the summary is exclusively focused on the identified reportable criteria. Para. [0022], the incident can be a resolved incident);
generating, with the first ( Li: Para.[0117],[0132], Fig. 5, The summarization tool 506 transmits a summarization request to a language model, such a summarization model 514, to obtain the status report. The summarization request includes instructions to the large language model to use the identified reportable criteria. The summarization tool 506 causes the summarization model 514 to extract a summary ( second narrative) from the incident information where the summary is exclusively focused on the identified reportable criteria. Para. [0022], the incident can be an unresolved incident);
and outputting the first narrative or the second narrative corresponding to each respective one of the one or more issue resolution indications( Li: Para.[0152], Fig. 10, at 1010 an incident status report ( summary) received from the language model is transmitted to one or more users (e.g. ,stakeholders).
Regarding Claim 20, Li in view of Ghoche teach the method of claim 19. Ghoche further teaches, further comprising:
generating, with the first large language model for the one or more issue resolution indications indicating resolved status, a first confidence score corresponding to one of: a probability that the resolved status is in fact indicating an issue, corresponding to the issue resolution indication indicating resolved status, is resolved, [or a probability that the unresolved status is in fact indicating an issue, corresponding to the issue resolution indication indicating unresolved status, is unresolved] ( Ghoche: Para.[0087]-[0089], a confidence can be generated for a particular answer ( resolved status) which was generated by a particular macro rather than from other macros. Para.[0088], BERT or other transformer based ML techniques can be used);
determining whether the first confidence score is greater than or equal to a threshold ( Ghoche: Para.[0087], a confidence ( first) can be generated for a particular answer ( resolved status) which was generated by a particular macro rather than from other macros. Para.[0088], BERT or other transformer based ML techniques can be used. Para.[0124], a template answer exceeds a selected threshold);
storing the issue resolution indication in one or more memories based on the determination that the first confidence score is greater than or equal to the threshold ( Ghoche: Para.[0266], storing in memory),
and generating, with a second first large language model based on the interaction transcript and the resolution prompt, one or more additional issue resolution indications and a second confidence score, based on the determination that the first confidence score is not greater than or equal to the threshold ( Ghoche: Para.[0114],generating confidence ( second) for partial completing a response answer. Para.[0124], the template answer doesn’t exceed threshold, the question of the ticket is routed to a human agent to resolve).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate Ghoche’s teaching of an autonomous customer support Chatbot Agent utilizing a large language model to aid in implementing a workflow to solve a customer issue, into the system and method, taught by Li, because, this would improve overall system performance by using an autonomous AI chatbot that interacts with a large language model to implement an improved workflow.( Ghoche, Para.[0199],[0221]).
Claims 11-13 are rejected under 35 U.S.C. 103 as being unpatentable over Li et al. ( US 20250005589 A1), hereinafter referenced as Li, in view of Ghoche et al. (US 20240386213 A1), hereinafter referenced as Ghoche, further in view of Haikin et al. ( US 20250165720 A1), hereinafter referenced as Haikin.
Regarding Claim 11, Li in view of Ghoche teach the method of claim 1. Li in view of Ghoche, fail to explicitly teach the claimed, further comprising: receiving a plurality of first narratives corresponding to a plurality of interactions; converting textual data structure of the plurality of first narratives into numerical vectors embeddings; discerning, with a categorization component processing the numerical vectors embeddings, one or more categories that are present within the plurality of first narratives; and labeling, with a label generation component, the one or more categories with a keyphrase.
However, Haikin does teach the claimed, further comprising: receiving a plurality of first narratives corresponding to a plurality of interactions ( Haikin: Para.[0063], answers or insights of interactions are generated by LLM);
converting textual data structure of the plurality of first narratives into numerical vectors embeddings ( Haikin: Para.[0064], Fig. 3, at 305, transforming the generated answers/insights into vector embeddings);
discerning, with a categorization component processing the numerical vectors embeddings, one or more categories that are present within the plurality of first narratives ( Haikin: Para.[0066], a clustering algorithm may be used to identify aggregations or clusters occurring within each. That is, with intents, sentiment-aspects, action items, etc. being stored as both unstructured fields and as vector embeddings, aggregations over each of the different insight types may be done via a clustering algorithm so to identify the similar clusters occurring within each);
and labeling, with a label generation component, the one or more categories with a keyphrase ( Haikin: Para.[0073], generating name for the clusters by the LLM, such as a first cluster would relate to “Shipping Issues”, while a second cluster would relate to “Product Quality”).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate Haikin’s teaching of natural language processing and related analytics in the field of customer service and customer relations management via contact centers, into the system and method, taught by Li in view of Ghoche, because, this would improve processing and analysis of natural language conversation data for enhancing search capabilities and related analytics.( Haikin, Para.[0001]-[0003]).
Regarding Claim 12, Li in view of Ghoche, further in view of Haikin teach the method of claim 11. Haikin further teaches, wherein the keyphrase generated by the label generation component is a phrase extracted from an overlapping portion of the plurality of first narratives categorized within each of the one or more categories ( Haikin: Para. [0068]-[0073], keyphrase “Product Quality” is generated by overlapping portion of the narratives of customer issues) ;
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate Haikin’s teaching of natural language processing and related analytics in the field of customer service and customer relations management via contact centers, into the system and method, taught by Li in view of Ghoche, because, this would improve processing and analysis of natural language conversation data for enhancing search capabilities and related analytics.( Haikin, Para.[0001]-[0003]).
Regarding Claim 13, Li in view of Ghoche, further in view of Haikin teach the method of claim 11. Haikin further teaches, wherein the keyphrase generated by the label generation component is an abstraction based on the plurality of first narratives categorized within each of the one or more categories ( Haikin: Para. [0068]-[0073], keyphrase “Shipping Issues” is generated by summarizing the narratives of customer issues) ;
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate Haikin’s teaching of natural language processing and related analytics in the field of customer service and customer relations management via contact centers, into the system and method, taught by Li in view of Ghoche, because, this would improve processing and analysis of natural language conversation data for enhancing search capabilities and related analytics.( Haikin, Para.[0001]-[0003]).
Claim 14 is rejected under 35 U.S.C. 103 as being unpatentable over Li et al. ( US 20250005589 A1), hereinafter referenced as Li, in view of Ghoche et al. (US 20240386213 A1), hereinafter referenced as Ghoche, further in view of Patil et al. ( US 20250278688 A1), hereinafter referenced as Patil.
Regarding Claim 14, Li in view of Ghoche teach the method of claim 1. Li in view of Ghoche fail to explicitly teach the claimed, further teaches, wherein obtaining the interaction transcript comprises: receiving an audio recording of the interaction between the first entity and the second entity, and generating, with a fourth large language model configured for speech recognition processing, the interaction transcript.
However, Patil does teach the claimed, further teaches, wherein obtaining the interaction transcript comprises:
receiving an audio recording of the interaction between the first entity and the second entity ( Patil: Para.[0044], interaction can be voice recording between the customer and agent) ,
and generating, with a fourth large language model configured for speech recognition processing, the interaction transcript ( Patil: Para.[0044], Non-text-based interactions may be converted into text-based interaction recordings (e.g., using automatic speech recognition). Para.[0054], transformer NN can be used for speech applications).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate Patil’s teaching of generating evaluation forms for the evaluation of agents in contact centers using interaction recordings, into the system and method, taught by Li in view of Ghoche, because, the automatic generation of evaluation forms may increase the accuracy of evaluation questions, and the service provided by an agent of a contact center to a customer can be improved as a result of better evaluation. ( Patil, Para.[0007]-[0010]).
Claims 23, 24 are rejected under 35 U.S.C. 103 as being unpatentable over Li et al. ( US 20250005589 A1), hereinafter referenced as Li, in view of Ghoche et al. (US 20240386213 A1), hereinafter referenced as Ghoche, further in view of Leary et al. ( US 20240095463 A1), hereinafter referenced as Leary.
Regarding claim 23, Li in view of Ghoche teach the method of claim 1. Li in view of Ghoche fail to explicitly teach the claimed, wherein the second large language model is different from the first large language model.
However, Leary does teach the claimed, wherein the second large language model is different from the first large language model ( Leary: Para.[0026], [0027], Fig. 1, the system architecture of a datacenter shows LLM inference service 102, where there are instances of the LLM at different sizes, a request may specify a model to use, such as a smaller model that is (at least with respect to one or more of compute, memory, or storage requirements) less expensive to host or operate, and takes less time to perform a computation, but is best for use with simple inferencing tasks, or a “large” model that may be more expensive to operate but provides more robust inferencing capability).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate Leary’s teaching of natural language processing applications using large language models, into the system and method, taught by Li in view of Ghoche, because an improved efficiency can be achieved by using LLMs of different sizes, such as with 1 billion, 5 billion, 40 billion, and 530 billion parameters, from small to large LLM instances .( Leary, Para.[0028]).
Regarding claim 24, Li in view of Ghoche teach the method of claim 1. Li in view of Ghoche fail to explicitly teach the claimed, wherein the first large language model comprises one of a small language model (SLM) or a low complexity large language model.
However, Leary does teach the claimed, wherein the first large language model comprises one of a small language model (SLM) or a low complexity large language model ( Leary: Para.[0035], Fig. 2, clients request can be directed to small LLM 214).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate Leary’s teaching of natural language processing applications using large language models, into the system and method, taught by Li in view of Ghoche, because an improved efficiency can be achieved by using LLMs of different sizes, such as with 1 billion, 5 billion, 40 billion, and 530 billion parameters, from small to large LLM instances .( Leary, Para.[0028]).
Allowable Subject Matter
Claims 3, 21 and 22 contain subject matter that is allowable over the prior art of record. They would be considered allowable if rewritten to include all of the limitations of the base claim and any intervening claims.
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 extension fee 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 date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to NADIRA SULTANA whose telephone number is (571)272-4048. The examiner can normally be reached M-F,7:30 am-5:00pm.
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, Paras D. Shah can be reached on (571) 270-1650. 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.
/NADIRA SULTANA/Examiner, Art Unit 2653
/Paras D Shah/Supervisory Patent Examiner, Art Unit 2653
05/28/2026
Prosecution Insights