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 .
Status of Claims
Applicant filed an amendment on May 26, 2026. Claims 1-20 were pending in the Application. Claims 1-2, 7-11, and 14-17 are amended. Claims 21-23 have been added. Claims 3, 13, and 18 have been canceled. Claims 1, 11, and 16 are the independent claims, the remaining claims depend on claims 1, 11, and 16. Thus claims 1-2, 4-12, 14-17, and 19-23 are currently pending. After careful and full consideration of Applicant arguments and amendments, the Examiner finds them to be moot and/or not persuasive.
Claim Objections
Claim 1 is objected to because of the following informalities: “A method to optimize a virtual agent leveraging a Large Language Model (LLM) model to determine …” should read “A method to optimize a virtual agent leveraging a Large Language Model (LLM) to determine …” . Additionally, similar language is recited in claims 11 and 16.
Claim 1 is objected to because of the following informalities: “providing, …, wherein the LLM model is fine-tuned …” should read “providing, …, wherein the LLM is fine-tuned …” . Additionally, similar language is recited in claims 11 and 16.
Claim 9 is objected to because of the following informalities: “The method of claim 1, … to the LLM model to generate…” should read “The method of claim 1, … to the LLM to generate…” . Additionally, similar language is recited in claims 10 and 14-15.
Claim 22 is objected to because of the following informalities: “The non-transitory, computer-readable medium of claim 11, wherein the method further comprises: retrieving …; performing …; generating …; performing …; and generating …” should read “The non-transitory, computer-readable medium of claim 11, wherein machine executable instructions being executed by the processor, further cause the processor to: retrieve …; perform …; generate …; perform …; and generate …” .
Response to Arguments
In the context of 35 U.S.C. §101, Applicant respectfully traverses the rejection in view of amended independent claims. Applicant is of the opinion that the claims are statutory and respectfully asserts that “amended independent claims 1, 11, and 16 are not directed to the alleged abstract idea, rather the claims are directed to a system that optimizes a virtual agent leveraging a Large Language Model (LLM) model to determine an outcome for an electronic transaction between a user and an online system which are completely technically in nature and cannot be categorized as commercial or legal interactions (i.e., organizing human activity), and an abstract idea; “routing the input …” involves processors, memory and network hardware which cannot be categorized as organizing a human activity; “modifying an aspect of virtual agent …”involves technical adjustment of software or adjusting dialogue flow which cannot be categorized as organizing human activity; the present disclosure also enables tailored control enabling bespoke customization and fine-tuning behavior of the LLM to specific needs and preferences; the “LLM model is fine-tune ...” involves adjustment in model parameters using datasets and generating outputs which cannot be categorized as commercial or legal interactions (i.e., organizing human activity), and does not direct to an abstract idea; the amended independent claim 1 integrates the alleged judicial exception into a practical application; the emphasized claimed technical features address the technical problem associated with the existing online system; the emphasized claimed features as a whole solves the technical problem by providing predictable outputs through calibration to deliver consistent responses (including the reliable information) that align with expected outcomes, reducing unexpected interactions with users; this includes consistency in the responses by maintaining a uniform response structure and preserving a conversational context; the present disclosure may also provide for efficiencies in terms of technical resource consumption, which also includes minimizing latency (even under heavy loads); the present disclosure also enable tailored control enabling bespoke customization and fine-tuning behavior of the LLM to specific needs and preferences; the amended features, according to the reminder guidelines of USPTO, cannot be performed in human mind, rather the above claimed steps provide flexibility in integration of the virtual agent into existing workflows and software systems (e.g., customer relationship management (CRM) systems, enterprise resource planning (ERP) systems); this includes handling structured data and unstructured data and seamless connection with existing databases, file systems, and the like to enable unified processes; the present disclosure also enhance scalability to scale in response to demand without compromising quality or performance; amended independent claim 1 integrates the purported judicial exception into the practical application as the amended independent claim 1 clearly identifies that "claim improves technology or a technical field"; the amended independent claim 1 "covers a particular solution to a problem or a particular way to achieve a desired outcome" as reminded in the guidelines; the claim recites a specific computing architecture and a process for optimizing a virtual agent leveraging an LLM model to determine an outcome; the claimed steps are specific technical mechanisms that constrain the implementation beyond "using a computer as a tool"; when considered as an ordered combination, amended claim 1 includes additional elements that amount to significantly more than any alleged judicial exception, and the § 101 rejection should be withdrawn; and claim 1, as amended, is not directed to the alleged abstract idea and is instead directed to patent eligible subject matter”.
Initially, the Examiner would like to point out that the basis of the rejection is Alice, by applying the subject matter eligibility analysis and flowchart according to MPEP § 2106, which applies a two-step framework, earlier set out in Mayo Collaborative Services v. Prometheus Laboratories, Inc., 566 U.S. 66 (2012), "for distinguishing patents that claim laws of nature, natural phenomena, and abstract ideas from those that claim patent-eligible applications of those concepts." Alice, 573 U.S. at 217.
Under the two-step framework, it must first be determined if "the claims at issue are directed to a patent-ineligible concept." If the claims are determined to be directed to a patent-ineligible concept, e.g., an abstract idea, then the second step of the framework is applied to determine if "the elements of the claim ... contain an "inventive concept" sufficient to 'transform' the claimed abstract idea into a patent-eligible application." (citing Mayo, 566 U.S. at 72-73, 79).
With regard to step one of the Alice framework, we apply a "directed to" two-prong test: 1) evaluate whether the claim recites a judicial exception, and 2) if the claim recites a judicial exception, evaluate whether the claim "applies, relies on, or uses the judicial exception in a manner that imposes a meaningful limit on the judicial exception, such that the claim is more than a drafting effort designed to monopolize the judicial exception," i.e., whether the claim integrates the judicial exception into a practical application. (MPEP §2106.04 II.A.1. and II.B.2.).
The Specification, para 1, provides evidence as to what the claimed invention is directed. In this case, the specification, para 1, discloses that the invention relates to determining an outcome for a transaction, and is grouped under “Certain Methods of Organizing Human Activity, commercial or legal interactions (including agreements in the form of contracts; legal obligations; advertising, marketing or sales activities or behaviors; business relations)”, and is grouped under “Certain Methods of Organizing Human Activity, managing personal behavior or relationships or interactions between people (including social activities, teaching, and following rules or instructions)” in prong one of step 2A. (MPEP §2106.04 II.A.1.).
Claim 1 provides additional evidence, and recites the limitations “receiving, by the one or more processors, input from the user of a user device, wherein the input includes request information or a query associated with the conversation; extracting and processing, by the one or processors, the request information from the input to generate state information, wherein the state information includes intent information, emotional state information, slot information, and constraint information; routing, by the one or more processors, the input and the generated state information by enabling the virtual agent to facilitate the conversation between the user and the online system; retrieving, by the one or more processors, information associated with the conversation from a Retrieval-Augmented Generation (RAG) database, wherein the information is associated with one or more of a product, service, or entity associated with the conversation; generating, by the one or more processors, a user profile for the user associated with the conversation, wherein the user profile is generated based on the generated state information, historical information, and demographic information associated with the user; modifying, by the one or more processors, an aspect of the virtual agent, configured to facilitate the conversation based on the information associated with the conversation and the user profile, by dynamically adapting operation of the virtual agent based on the information associated with the conversation and the user profile; communicating, by the one or more processors, an initial proposition to the user via an initial proposition communication, wherein the initial proposition is based on an initial engagement strategy to facilitate the conversation; interacting, by the one or more processors, via the virtual agent, with the user to obtain additional information associated with the conversation and the initial proposition; dynamically modifying, by the one or more processors, the initial proposition to generate an alternative proposition, including generating an alternative engagement strategy to facilitate the conversation based on the initial engagement strategy and the additional information, wherein the alternative engagement strategy is generated based on adjusting of weighting values by evaluating a plurality of engagement strategies; communicating, by the one or more processors, the alternative proposition to the user via an alternative proposition communication; receiving, by the one or more processors, via the virtual agent, a response to the alternative proposition from the user; determining, by the one or more processors, the outcome for the conversation based on the response to the alternative proposition; providing, by the one or more processors, via the virtual agent, the outcome for the conversation to the user, wherein the LLM model is fine-tuned by generating the outcome using the state information and the user profile”, which represent the abstract idea of “a transaction outcome”. The abstract idea is in italics, and the additional elements are in bold. (MPEP §2106.04 II.A.1.).
This judicial exception is not integrated into a practical application because, when analyzed under prong two of step 2A (MPEP §2106.04 II.A.2.), the additional elements of the claim, such as “a virtual agent leveraging a Large Language Model (LLM) model”, “a conversation”, “an online system over a computer network using, one or more processors”, “a user device”, “the virtual agent to facilitate the conversation”, “the conversation from a Retrieval-Augmented Generation (RAG) database”, and “interacting, by the one or more processors, via the virtual agent”, amount to merely “apply it”, as they represent the use of a computer as a tool to perform an abstract idea. Therefore, the additional elements do not integrate the abstract idea into a practical application as they do no more than represent a computer performing functions that correspond to implementing the acts of “a transaction outcome”.
Examiner notes the basis of the rejection was, and is not as any mental process covering performance in the mind, but classified as an abstract idea, “a transaction outcome”, and is grouped under “Certain Methods of Organizing Human Activity, commercial or legal interactions (including agreements in the form of contracts; legal obligations; advertising, marketing or sales activities or behaviors; business relations)” , and is grouped under “Certain Methods of Organizing Human Activity, managing personal behavior or relationships or interactions between people (including social activities, teaching, and following rules or instructions)”.
With respect to the additional elements operating in a non-conventional and non-generic way and reflecting an improvement to a particular technological environment, the cited additional elements represent the use of a computer as a tool to perform an abstract idea. Therefore, the additional elements do not integrate the abstract idea into a practical application as they do no more than represent a computer performing functions that correspond to implementing the acts of “a transaction outcome.” The claim is not directed to improving computer functionality nor improving another technology or technical field, but improving the method for “a transaction outcome”. For potential improvement in an abstract idea, “a transaction outcome”, it is important to keep in mind that an improvement in the abstract idea itself (e.g. a transaction outcome concept) is not an improvement in technology. (MPEP § 2106.04(d)(1)). Therefore, claim 1 is non-statutory.
Claim 11 also recites the abstract idea of “a transaction outcome”, as well as the additional elements of “a non-transitory, computer-readable medium including machine-readable instructions that are executable by a processor to optimize a virtual agent leveraging a Large Language Model (LLM) model to determine an outcome for a conversation between a user and an online system over a computer network to: …”, “a user device”, “the virtual agent to facilitate the conversation ”, “the conversation from a Retrieval-Augmented Generation (RAG) database”, “and “interact, via the virtual agent”, which amount to merely “apply it”, as they represent the use of a computer as a tool to perform an abstract idea. Therefore, the additional elements do not integrate the abstract idea into a practical application as they do no more than represent a computer performing functions that correspond to implementing the acts of “a transaction outcome”.
When analyzed under step 2B (MPEP 2106.05 I.A.), the claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception itself. Viewed as a whole, the combination of elements recited in the claim merely describe the concept of “a transaction outcome” using computer technology (e.g., “a non-transitory, computer-readable medium” and “a Retrieval-Augmented Generation (RAG) database”). Therefore, the use of these additional elements do no more than employ a computer as a tool to implement the abstract idea. And as the computer does no more than serve as a tool to implement the abstract idea, they do not improve computer functionality nor improve another technology or technical field. Therefore, claim 11 is non-statutory.
Claim 16 also recites the abstract idea of “a transaction outcome”, as well as the additional elements of “a system to optimize a virtual agent leveraging a Large Language Model (LLM) model to determine an outcome for a conversation between a user and an online system over a computer network using, one or more processors, comprising: …”, “a processor”, “a non-transitory memory device including machine-readable instructions that are executable by the processor to: …”, “a user device”, “the virtual agent to facilitate the conversation ”, “the conversation from a Retrieval-Augmented Generation (RAG) database”, “and “interact, via the virtual agent”, which amount to merely “apply it”, as they represent the use of a computer as a tool to perform an abstract idea. Therefore, the additional elements do not integrate the abstract idea into a practical application as they do no more than represent a computer performing functions that correspond to implementing the acts of “a transaction outcome”.
When analyzed under step 2B (MPEP 2106.05 I.A.), the claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception itself. Viewed as a whole, the combination of elements recited in the claim merely describe the concept of “a transaction outcome” using computer technology (e.g., “a processor” and “a non-transitory memory device”). Therefore, the use of these additional elements do no more than employ a computer as a tool to implement the abstract idea. And as the computer does no more than serve as a tool to implement the abstract idea, they do not improve computer functionality nor improve another technology or technical field. Therefore, claim 16 is non-statutory.
Finally, Examiner notes the basis of the rejection is Alice, by applying the subject matter eligibility analysis and flowchart according to MPEP § 2106. And, based on this standard, the claims are non-statutory, and correctly rejected under 35 U.S.C. § 101.
In the context of 35 U.S.C. § 112(b), Means-Plus-Function, paragraph 11 of the Non-Final Rejection Office Action dated March 26, 2026, Applicant has amended claims 1, 11, and 16 to render the rejection under 35 U.S.C. § 112(b), Means-Plus-Function, moot. Claims 1, 11, and 16 have been amended to eliminate the claim language “a strategic selection model configured to evaluate”, which was the basis for the rejection under 35 U.S.C. § 112(b), Means-Plus-Function. Examiner hereby rescinds the rejection under 35 U.S.C. § 112(b), Means-Plus-Function, paragraph 11 of the Non-Final Rejection Office Action dated March 26, 2026.
In the context of 35 U.S.C. § 103, in the Non-Final Rejection Office Action dated March 26, 2026, Applicant submits that none of the references Ben David or Mico individually or in combination teach or suggest the instant recitation of amended independent claim 1. Further, one of ordinary skills in art would have found no motivation to combine and modify these references to arrive at the features of amended claim 1.
Dependent claims depending on independent claims 1, 11, and 16 are patentably distinct for at least the reasons discussed with respect to amended independent claims 1, 11, and 16.
Further, withdrawal of the rejections and allowance of claims 2, 4-10, 12, 14-15, 17, and 19-20 are therefore respectfully requested.
Therefore, for the above reasons the rejection under 35 U.S.C. 103 is respectfully traversed, and withdrawal of the rejection is respectfully requested, or in any combination, do not teach or suggest the features of the amended claims.
Examiner finds the applicant arguments for the limitations “receiving, by the one or more processors, input from the user of a user device, wherein the input includes request information or a query associated with the conversation”;
“extracting and processing, by the one or processors, the request information from the input to generate state information, wherein the state information includes intent information, emotional state information, slot information, and constraint information”;
“routing, by the one or more processors, the input and the generated state information by enabling the virtual agent to facilitate the conversation between the user and the online system”;
“retrieving, by the one or more processors, information associated with the conversation from a Retrieval-Augmented Generation (RAG) database, wherein the information is associated with one or more of a product, service, or entity associated with the conversation”;
“… wherein the user profile is generated based on the generated state information, historical information, and demographic information associated with the user”;
“… by dynamically adapting operation of the virtual agent based on the information associated with the conversation and the user profile”;
“providing, by the one or more processors, via the virtual agent, the outcome for the conversation to the user, wherein the LLM model is fine-tuned by generating the outcome using the state information and the user profile”, moot in view of new grounds of rejection, and therefore, amended claim 1, as well as amended claims 11 and 16, is not patentable. Amended claim 1, as well as amended claims 11 and 16, stands rejected under 35 U.S.C §103 in the analysis below, and is therefore, not patentable in view of D’Agostino (US 20250307834 A1) and Bathwal (US 12314318 B2) now applying to the applicable amended sections for claim 1, as well as to amended claims 11 and 16. Dependent claims 2, 4-10, and 21, which depend on claim 1; dependent claims 12, 14-15, and 22, which depend on claim 11; and dependent claims 17, 19-20, and 23, which depend on claim 16, also stand rejected under 35 U.S.C. § 103.
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, 4-10, 11-12, 14-15, 16-17, and 19-23 are rejected under 35 U.S.C. § 101 because the claimed invention is directed to an abstract idea without significantly more.
In the instant case, claims 1-2, 4-10, and 21 are directed to a “method”; claims 11-12, 14-15, and 22 are directed to a “non-transitory, computer readable medium”; and claims 16-17, 19-20, and 23 are directed to a “system”. Therefore, these claims are directed to one of the four statutory categories of invention.
Claim 1 recites “a transaction outcome”, which is a form of commercial or legal interactions and is a form of managing personal behavior or relationships or interactions between people (i.e., organizing human activity), and an abstract idea. Specifically, the claim recites “receiving, by the one or more processors, input from the user of a user device, wherein the input includes request information or a query associated with the conversation; extracting and processing, by the one or processors, the request information from the input to generate state information, wherein the state information includes intent information, emotional state information, slot information, and constraint information; routing, by the one or more processors, the input and the generated state information by enabling the virtual agent to facilitate the conversation between the user and the online system; retrieving, by the one or more processors, information associated with the conversation from a Retrieval-Augmented Generation (RAG) database, wherein the information is associated with one or more of a product, service, or entity associated with the conversation; generating, by the one or more processors, a user profile for the user associated with the conversation, wherein the user profile is generated based on the generated state information, historical information, and demographic information associated with the user; modifying, by the one or more processors, an aspect of the virtual agent, configured to facilitate the conversation based on the information associated with the conversation and the user profile, by dynamically adapting operation of the virtual agent based on the information associated with the conversation and the user profile; communicating, by the one or more processors, an initial proposition to the user via an initial proposition communication, wherein the initial proposition is based on an initial engagement strategy to facilitate the conversation; interacting, by the one or more processors, via the virtual agent, with the user to obtain additional information associated with the conversation and the initial proposition; dynamically modifying, by the one or more processors, the initial proposition to generate an alternative proposition, including generating an alternative engagement strategy to facilitate the conversation based on the initial engagement strategy and the additional information, wherein the alternative engagement strategy is generated based on adjusting of weighting values by evaluating a plurality of engagement strategies; communicating, by the one or more processors, the alternative proposition to the user via an alternative proposition communication; receiving, by the one or more processors, via the virtual agent, a response to the alternative proposition from the user; determining, by the one or more processors, the outcome for the conversation based on the response to the alternative proposition; providing, by the one or more processors, via the virtual agent, the outcome for the conversation to the user, wherein the LLM model is fine-tuned by generating the outcome using the state information and the user profile”. The abstract idea is in italics, and the additional elements are in bold. (MPEP §2106.04 II.A.1.).
This judicial exception is not integrated into a practical application because, when analyzed under prong two of step 2A (MPEP §2106.04 II.A.2.), the additional elements of the claim, such as “a virtual agent leveraging a Large Language Model (LLM) model”, “a conversation”, “an online system over a computer network using, one or more processors”, “a user device”, “the virtual agent to facilitate the conversation”, “the conversation from a Retrieval-Augmented Generation (RAG) database”, and “interacting, by the one or more processors, via the virtual agent”, amount to merely “apply it”, as they represent the use of a computer as a tool to perform an abstract idea. Therefore, the additional elements do not integrate the abstract idea into a practical application as they do no more than represent a computer performing functions that correspond to implementing the acts of “a transaction outcome”.
When analyzed under step 2B (MPEP 2106.05 I.A.), the claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception itself. Viewed as a whole, the combination of elements recited in the claim merely describes the concept of “a transaction outcome” using computer technology (e.g., “one or more processors” and “an online system”). Therefore, these additional elements do no more than employ a computer as a tool to implement the abstract idea. And as the computer does no more than serve as a tool to implement the abstract idea, they do not improve computer functionality nor improve another technology or technical field. Therefore, claim 1 is non-statutory.
Claim 11 also recites the abstract idea of “a transaction outcome”, as well the additional elements of “a non-transitory, computer-readable medium including machine-readable instructions that are executable by a processor to optimize a virtual agent leveraging a Large Language Model (LLM) model to determine an outcome for a conversation between a user and an online system over a computer network to: …”, “a user device”, “the virtual agent to facilitate the conversation ”, “the conversation from a Retrieval-Augmented Generation (RAG) database”, “and “interact, via the virtual agent”, which amount to merely “apply it”, as they represent the use of a computer as a tool to perform an abstract idea. Therefore, the additional elements do not integrate the abstract idea into a practical application as they do no more than represent a computer performing functions that correspond to implementing the acts of “a transaction outcome”.
When analyzed under step 2B (MPEP 2106.05 I.A.), the claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception itself. Viewed as a whole, the combination of elements recited in the claim merely describes the concept of “a transaction outcome” using computer technology (e.g., “a non-transitory, computer-readable medium” and “a Retrieval-Augmented Generation (RAG) database”). Therefore, these additional elements do no more than employ a computer as a tool to implement the abstract idea. And as the computer does no more than serve as a tool to implement the abstract idea, they do not improve computer functionality nor improve another technology or technical field. Therefore, claim 11 is non-statutory.
Claim 16 also recites the abstract idea of “a transaction outcome”, as well the additional elements of “a system to optimize a virtual agent leveraging a Large Language Model (LLM) model to determine an outcome for a conversation between a user and an online system over a computer network using, one or more processors, comprising: …”, “a processor”, “a non-transitory memory device including machine-readable instructions that are executable by the processor to: …”, “a user device”, “the virtual agent to facilitate the conversation ”, “the conversation from a Retrieval-Augmented Generation (RAG) database”, “and “interact, via the virtual agent”, which amount to merely “apply it”, as they represent the use of a computer as a tool to perform an abstract idea. Therefore, the additional elements do not integrate the abstract idea into a practical application as they do no more than represent a computer performing functions that correspond to implementing the acts of “a transaction outcome”.
When analyzed under step 2B (MPEP 2106.05 I.A.), the claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception itself. Viewed as a whole, the combination of elements recited in the claim merely describes the concept of “a transaction outcome” using computer technology (e.g., “a non-transitory memory device” and “a processor”). Therefore, these additional elements do no more than employ a computer as a tool to implement the abstract idea. And as the computer does no more than serve as a tool to implement the abstract idea, they do not improve computer functionality nor improve another technology or technical field. Therefore, claim 16 is non-statutory.
Dependent claims 2, 4-10, 12, 14-15, 17, and 19-23 further describe the abstract idea of “a transaction outcome”, which is insufficient to overcome the rejections of claims 1, 11, and 16.
Dependent claims 4-10, 12, 14-15, and 19-20 do not recite any new additional elements that integrate the abstract idea into a practical application, and that do no more than represent a computer performing functions that correspond to implementing the acts of “a transaction outcome”, when analyzed under Step 2A, Prong Two. And, as they do no more than employ a computer as a tool to implement the abstract idea, they do not improve computer functionality nor improve another technology or a technical field, when analyzed under Step 2B.
Dependent claims 2 and 17 recite a new additional element of “a future conversation”, which does no more than employ a computer as a tool to implement the abstract idea. And, as it does no more than employ a computer as a tool to implement the abstract idea, it does not improve computer functionality nor improve another technology or a technical field.
Dependent claims 21-23 recite a new additional element of “an internal database”, which does no more than employ a computer as a tool to implement the abstract idea. And, as it does no more than employ a computer as a tool to implement the abstract idea, it does not improve computer functionality nor improve another technology or a technical field.
Hence, claims 1-2, 4-10, 11-12, 14-15, 16-17, and 19-23 are not patent eligible.
Claim Rejections - 35 USC § 112
The following is a quotation of 35 U.S.C. 112(d):
(d) REFERENCE IN DEPENDENT FORMS.—Subject to subsection (e), a claim in dependent form shall contain a reference to a claim previously set forth and then specify a further limitation of the subject matter claimed. A claim in dependent form shall be construed to incorporate by reference all the limitations of the claim to which it refers.
The following is a quotation of pre-AIA 35 U.S.C. 112, fourth paragraph:
Subject to the following paragraph [i.e., the fifth paragraph of pre-AIA 35 U.S.C. 112], a claim in dependent form shall contain a reference to a claim previously set forth and then specify a further limitation of the subject matter claimed. A claim in dependent form shall be construed to incorporate by reference all the limitations of the claim to which it refers.
Claim 19 is rejected under 35 U.S.C. 112(d) or pre-AIA 35 U.S.C. 112, 4th paragraph, as being of improper dependent form for failing to further limit the subject matter of the claim upon which it depends, or for failing to include all the limitations of the claim upon which it depends.
Improper Dependent Claim
Claim 19 recites “The system of claim 18, wherein the alternative engagement strategy ...” A claim in dependent form shall be construed to incorporate by reference all the limitations of the claim to which it refers. Following the statute, the test as to whether a claim is a proper dependent claim is that it shall include every limitation of the claim from which it depends and specify a further limitation of the subject matter claimed. Claim 18 has been canceled, thus, claim 19 cannot include every limitation of claim 18, and therefore, does not specify a further limitation of the canceled claim 18. Therefore, claim 19 is not a proper dependent claim. For examination purposes, claim 19 is being interpreted as depending from claim 16.
Applicant may cancel the claim(s), amend the claim(s) to place the claim(s) in proper dependent form, rewrite the claim(s) in independent form, or present a sufficient showing that the dependent claim(s) complies with the statutory requirements.
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 set forth in Graham v. John Deere Co., 383 U. S. 1. 148 USPQ 459 (1966), that are applied for establishing a background for determining obviousness under 35 U.S.C. § 103 are summarized as follows:
Determining the scope and contents of the prior art.
Ascertaining the differences between the prior art and the claims at issue.
Resolving the level of ordinary skill in the pertinent art.
Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claims 1-2, 4-12, 14-17, and 19-23 are rejected under 35 U.S.C. 103 as being unpatentable over Ben David et al (U. S. Patent No. 11875123 B1), herein referred to as Ben David, in view of Mico et al (U. S. Patent Application Publication No. 20240338710 A1), herein referred to as Mico, in view of D’Agostino (U. S. Patent Application Publication No. 20250307834 A1), herein referred to as D’Agostino, and in further view of Bathwal et al (U. S. Patent No. 12314318 B2), herein referred to as Bathwal.
Regarding claims 1, 11, and 16, Ben David discloses a method … comprising:
… wherein the information is associated with one or more of a product, service, or entity associated with the conversation (C/L 3/21-29, “… Relevant data is gathered from the user, including their current financial state, financial reports, and necessary transactional data. This data is input into the LLM and forms the basis for the personalized advice. Output from the LLM is analyzed by a classification model to characterize the user's 25 intent within a known ontology, enabling the system to address the specific problems raised by the user and offer targeted advice that is tailored to the customer's unique circumstances and requirements …”; C/L 4/19-24, “… Data repository (100) stores account(s) (110). Each account(s) (110) is a set of data about a user. The specific information included in the account(s) (110) would depend on the context and purpose However, the account(s) includes relevant details and characteristics of the user within a related system or platform …”);
generating, by the one or more processors (FIG. 1, item 112; C/L 4/34-44, “… The system shown in FIG. 1 includes one or more server(s) (112). The server(s) (112) is one or more computers, possibly communicating in a distributed computing environment. The server(s) (112) may include multiple physical and virtual computing systems that form part of a cloud computing environment. Thus, the server(s) (112) includes one or more processors. The processor can be hardware, or a virtual machine programmed to execute one or more controllers and/or software applications … the processors of the server(s) (112) may be the computer processor(s) (1002) of FIG. l0A …”; FIG. 10A, item 1002; C/L 13/53-57, “… FIG. l0A, the computing system (1000) may include one or more computer processor(s) (1002), …”; C/L 13/60-67, “…The computer processor(s) (1002) may be an integrated circuit for processing instructions. The computer processor(s) may be one or more cores or micro-cores of a processor. The computer processor(s) (1002) includes one or more processors. The one or more processors may include a central processing unit (CPU), a 65 graphics processing unit (GPU), a tensor processing unit (TPU), combinations thereof, etc. …”), a user profile for the user associated with the conversation, wherein the user profile is generated based on the generated state information, historical information, and demographic information associated with the user (C/L 4/25-33, “… account(s) (110) may include profile information such as the user's personal information, contact details, preferences, as well as any other demographic or identification data associated with their account. in the context of financial advice or planning, the account may financial data, such as user's financial records, transaction history, income, expenses, assets, account balances, liabilities, historical data, usage patterns, and/or other financial metrics …”); …
communicating, by the one or more processors (FIG. 1, item 112; C/L 4/34-44; FIG. 10A, item 1002; C/L 13/53-57; C/L 13/60-67), an initial proposition to the user via an initial proposition communication, wherein the initial proposition is based on an initial engagement strategy to facilitate the conversation (FIG. 3, items 350, 360; C/L 8/30-43, “… The plan can be formatted according to a prompt template, influencing the LLM to interpret inputs in an equivalent manner, and produce similarly formatted output. Upon receiving the plan as input, the LLM employs its natural language processing capabilities to analyze and understand the content of the plan, and to generate advice. The advice is generated in a natural language format. At Step 360, the advice is forwarded to the user device. The advice can be communicated directly to the user, device, or sent to the advice planner where it is forwarded to the user device. The advice may be transmitted back to the user device using appropriate communication protocols or APIs, ensuring the secure and efficient delivery of the advice …”);
interacting, by the one or more processors (FIG. 1, item 112; C/L 4/34-44; FIG. 10A, item 1002; C/L 13/53-57; C/L 13/60-67), via the virtual agent (FIG. 1, item 116; C/L 5/23-26, “… The server(s) (112) includes an advice planner (116). The advice planner (116) is one or more applications or application specific hardware programmed to control operation of the machine learning model(s) (118)…”; FIG. 4, item 410, C/L 8/54-64, “… At Step 410, a prompt to the LLM is generated from a template and the intent. To generate the prompt, the advice planner interacts with a prompt template and the intent. The prompt template provides a structured framework that guides the generation of the prompt. It may include placeholders or variables that are replaced with specific information based on the intent. The advice planner combines the prompt template and the intent to form a cohesive prompt that captures the necessary information for the LLM. This prompt is then provided as input to the LLM. The prompt serves as an input to the LLM …”; FIG. 9, item 914; C/L 12/47-61, “… The advice planner (914) is an example of the advice planner (116) of FIG. 1. The advice planner (914) receives this intent from intent/goal extraction (912) as the first input. Simultaneously, the advice planner (914) retrieves the current state (916) of the user's financial accounts, including their income, expenses, savings, investments, and debts. This state (916) of the user's account serves as the second input to the advice planner, providing crucial context and relevant financial information. The advice planner classifies the user's intent into a specific financial domain. This classification enables the planner to identify the appropriate context and framework for generating tailored financial advice … if the user's intent is focused on retirement planning, the advice planner categorizes the intent within the retirement domain …”), with the user to obtain additional information associated with the conversation and the initial proposition (C/L 3/4-9, “… Users can interact with the system by either asking their own questions or automatically generat-ing commonly asked and useful advice-related queries. Through continuous refinement and evaluation, the system can deliver valuable and actionable guidance tailored to individual needs of each user…”; C/L 7/3-6, “… Users can conveniently express their intents or desires by simply typing their request in natural language, making the interaction more intuitive and user-friendly…”; C/L 12/17-25, “… The user interacts with a financial planning application through their device, such as a smartphone or computer. The application is equipped with an advice planner, which serves as the core component responsible for generating personalized financial advice, the application be customer facing, such as an interface that allows a user to directly access and receive advice from the financial planning system … the application can be agent facing, where the user can a service agent providing live help to a customer …”);
dynamically modifying, by the one or more processors (FIG. 1, item 112; C/L 4/34-44; FIG. 10A, item 1002; C/L 13/53-57; C/L 13/60-67), the initial proposition to generate an alternative proposition, including generating an alternative engagement strategy to facilitate the conversation based on the initial engagement strategy and the additional information, wherein the alternative engagement strategy is generated based on adjusting of weighting values by evaluating a plurality of engagement strategies (FIG. 5, 6, items 510, 520, 530, 610, 620; C/L 9/44-54, “… At Step 510, a second state of the account is received … the advice planner receives the second state of the account at a later point in time … the example of providing financial advice, the second state represents an updated financial state. This updated state may be obtained through regular updates or interactions with the user's account, such as new transactions, changes in balances, or other relevant updates …”; C/L 9/52-54, 59-64, “ … At Step 520, method determines whether a first action logic of the plan was performed between the first state and the second state … If the first action logic was indeed performed, the advice planner proceeds with further adjustments. At Step 530, a set of weights associated with the first action logic is adjusted based on a difference between the first state and the second state …”; C/L 10/33-35, 44-51, “ … At Step 610, the set of weights is increased in response to determining that the first action logic was performed between the first state and the second state … At Step 620, the set of weights is decreased in response to determining that the first action logic was not perform between the first state and the second state …, if it is determined that the first action logic was not performed, suggesting a deviation from the desired outcome or an alternative approach, the set of weights is decreased. This adjustment reduces the impact of the first action logic, signaling the need for alternative actions or adjustments …”); …
In regards to claim 11, Ben David further discloses a non-transitory, computer-readable medium including machine-readable instructions that are executable by a processor (C/L 2/11-14 “… The one or more embodiments provide for a non-transitory computer readable storage medium storing program code which, when executed by a processor, performs a computer-implemented method …”; C/L 14/28-39, “ … Software instructions in the form of computer readable program code to perform embodiments may be stored, in whole or in part, temporarily or permanently, on a non-transitory computer readable medium such as a CD, DVD, storage device, a diskette, a tape, flash memory, physical memory, or any other computer readable storage medium. Specifically, the software instructions may correspond to computer readable program code that, when executed by a processor(s), is configured to perform one or more embodi-ments of the invention, which may include transmitting, receiving, presenting, and displaying data and messages described in the other figures of the disclosure…”) … to: …
In regards to claim 16, Ben David further discloses a system (FIG. 1; C/L 4/34-37, “… The system shown in FIG. 1 includes one or more server(s) (112). The server(s) (112) is one or more computers, possibly communicating in a distributed computing environment …”) … comprising:
a processor (FIG. 1, 10A, items 112, 1002; C/L 4/40-44, “… the server(s) (112) includes one or more processors. The processor can be hardware, or a virtual machine programmed to execute one or more controllers and/or software applications … the processors of the server(s) (112) may be the computer processor(s) (1002) of FIG. l0A …”; C/L 13/53-57, 60-67, “… FIG. l0A, the computing system (1000) may include one or more computer processor(s) (1002), non-persistent storage (1004), persistent storage (1006), a communication interface (1012) … The computer processor(s) (1002) may be an integrated circuit for processing instruc-tions. The computer processor(s) may be one or more cores or micro-cores of a processor. The computer processor(s) (1002) includes one or more processors. The one or more processors may include a central processing unit (CPU), a graphics processing unit (GPU), a tensor processing unit (TPU), combinations thereof, etc. …”);
a non-transitory memory device including machine-readable instructions that are executable by the processor (C/L 2/11-14 “… The one or more embodiments provide for a non-transi-tory computer readable storage medium storing program code which, when executed by a processor, performs a computer-implemented method …”; C/L 14/28-39, “ … Software instructions in the form of computer readable program code to perform embodiments may be stored, in whole or in part, temporarily or permanently, on a non-transitory computer readable medium such as a CD, DVD, storage device, a diskette, a tape, flash memory, physical memory, or any other computer readable storage medium. Specifically, the software instructions may correspond to computer readable program code that, when executed by a processor(s), is configured to perform one or more embodi-ments of the invention, which may include transmitting, receiving, presenting, and displaying data and messages described in the other figures of the disclosure…”) to: …
Ben David does not specifically disclose, however, Mico discloses modifying, by the one or more processors (FIG. 1, item 102; para 49, “FIG. 1 illustrates an automated chatbot server 102, …”; para 50, “… The automated chatbot server 102 may be a computing device and/or a software program that provides functionality for other programs (e.g., order management system 114) and/or devices integrated with a conversational software application (e.g., dynamic chatbot 104) of an enterprise to manage response 130 to a customer request 106 over a network 205. The automated chatbot server 102 may accept and/or respond to user requests 106 made over a network 205 by managing and sharing critical organization resources. The automated chatbot server 102 may be programmed to primarily execute processing of customer requests (e.g., user request 106). …”), an aspect of the virtual agent configured to facilitate the conversation based on the information associated with the conversation and the user profile (para 23, “The artificial intelligence based configurator of the disclosed system integrated within the conversational software application may include an intelligent indicator generation algorithm that may automatically change the selection and/or the order of button-based options within the chatbot. The system may explicitly provide recommendations to predict the most likely indicator options that will lead to satisfactory response to a customer's concerns and a successful outcome for the customer …”; para 25, “The disclosed system may include a process to automatically change the selection and/or the order of indicator-based options within the chatbot. The indicator generation algorithm … create and trigger options of smart indicators based on number of clicks counted across other users (e.g., interacting with the par-ticular enterprise and/or a wider group), past history of a particular user (e.g., page he/she is on, past order history, whether he/she has clicked on the "Track" button before, etc.), last clicked, machine learning recommendations such as from similar users with similar sentiment of message (e.g., using collaborative filtering or deep learning), external sources such as an API or database, user experience research methodology (e.g., A/B tests), predict what they're most likely trying to do based on any of the above factors and prioritize, reorder, and/or otherwise highlight choices (e.g., smart indicator) based on the interactional data …”), by dynamically adapting operation of the virtual agent based on the information associated with the conversation and the user profile (para 57, “The automated chatbot server 102 may be able to dynamically generate more natural and human-like responses, and it can also adapt and improve its responses (e.g., response 130) over time using machine learning methods and artificial intelligence to train its data resources …”; para 59, “… The dynamic chatbot 104 may use machine learning and natural language processing (NLP) techniques to generate more natural and human-like responses, and may adapt to the context and evolve over time. The disclosed dynamic chatbot 104 may boost engagement with a user 210 by generating automated messages, text, smart buttons (e.g., smart indicator 208), and/or images when interacting with the customers. The response generation module 206 of the automated chatbot server 102 may learn from past interactions and improve its responses 130, it may change its behavior based on the context of the conversation, it may generate new responses 130, and it may understand the intent behind the user input (e.g., user request 106) …”); …
communicating, by the one or more processors (FIG. 1, item 102; para 49; para 50), the alternative proposition to the user via an alternative proposition communication (para 29, “… the disclosed system may dynamically generate smart indicators based on a collection of data collected from the enterprise system (e.g., an order management system), machine learning, counting clicks, and what people have done in the past in similar situations, etc. The disclosed system may configure an ability to provide a better selection of the smart indicator that is more likely to achieve a faster, more intuitive response for the customer …”);
receiving, by the one or more processors (FIG. 1, item 102; para 49; para 50), via the virtual agent (para 35, “… The buyers and sellers may communicate using the disclosed framework in an optimized way that allows various aspects of decisions (e.g., purchasing decisions) to be negotiated and adjusted nearly instantly … the consumers use their client devices (e.g., laptop, smart phone, tablet, etc.) to anonymously organize and set up automated, on-going recommendation plans executed by multi-criteria decision-making negotiation agents on a server, referred to herein as "buyer AI negotiators" ("AI buyers," "AI buyer agent," "consumer bot," or "purchasing bot" for short), for purchasing particular products or services, either online or through physical kiosks or storefronts ... allow retailers to set up automated, on-going recommendation terminals or selling campaigns executed by multi-criteria decision-making negotiation agents on a server, referred to herein as "seller AI negotiators" (or "AI sellers" "AI seller agent" for short), that offer the seller's products or services and automatically recommendation with the buyer AI negotiators. The various embodiments may operate by combining large data sets of customer and consumer data from internal and/or external sources within an organization and applying intelligent, iterative processing algorithms to learn from patterns and features in the data that they analyze …”; para 50) a response to the alternative proposition from the user (FIG. 3, items 302, 350; para 74, “… a user may start an interactional event 305 with the automated chatbot server 102 through its dynamic chatbot 104. The automated chatbot server 102 may receive 302 a user request 106 through this interactional event 305…”); and
determining, by the one or more processors (FIG. 1, item 102; para 49; para 50), the outcome for the conversation based on the response to the alternative proposition (FIG. 3, items 102, 108, 110, 132, 204, 208, 304, 306, 308, 310, 350; para 74, “… The automated chatbot server 102 may use natural language processing and machine learning methods to analyze 304 the user request 106. The user message analysis module 108 of the automated chatbot server 102 may identify the user intent 110 and entity 306 from the ongoing interactional event 305 and user request 106. The processing module 124 of the automated chatbot server 102 may map 308 the contextual data 132 with the user intent 110 and the derived context information 112. Based on its real time mapping of context information 112, the response generation module 206 may look up at the query index 204 for matching context and/or situation. Once mapped, the response generation module 206 may present a smart indicator 208 found for similar context and situation … the response generation module 206 may trigger the artificial intelligence based configurator 126 to dynamically generate 310 a relevant set of indicators when the query index 204 is not mapped for matching context and/or situation…”); and …
Mico discloses customer assistance using AI-generated smart indicators. It would have been obvious to one of ordinary skill in the art before the effective filing date of the invention to include customer assistance using AI-generated smart indicators, as in Mico, to improve and/or enhance the technology of an advice generation system, as in Ben David, because it would amount to combining elements that in the combination would perform the same function as they functioned separately. One of ordinary skill in the art before the effective filing date of the invention would have been motivated to combine the references to provide the hardware and software for customer assistance using artificial intelligence (AI)-generated smart indicators, which may predict the most likely context-based indicator options that will lead to a successful outcome for the customer.
Ben David and Mico do not specifically disclose, however, D’Agostino discloses … to optimize a virtual agent leveraging a Large Language Model (LLM) model to determine an outcome for a conversation between a user and an online system (para 167, “… an apparatus includes functionality to facilitate continuous improvement and refinement of the apparatus's performance based on feedback received during communication sessions. The processor, coupled with memory and responsible for executing the LLM with multiple attention heads, is configured to receive feedback about the response generated during the communication session from one or more source devices and the service provider device. Upon receiving feedback, the processor leverages the feedback to retrain the LLM based on a combination of the response and the feedback about the response. The retraining process allows the apparatus to adapt and refine its response generation capabilities over time, improving the relevance and effectiveness of the generated responses in subsequent communication sessions … the processor utilizes the feedback to update the attention weight parameters within the LLM, optimizing its performance based on the specific needs and preferences of the users. By incorporating feedback-driven retraining into its operation, the apparatus enhances its ability to learn from user interactions and adjust its response generation strategies accordingly, leading to more personalized and satisfying user experiences …”; para 189, “… an apparatus optimizes future communication sessions based on insights gained from ongoing interactions …) over a computer network (FIG. 10, items 1050, 1060; para 214, “… accessing a network 1060 …”; para 221, “… Network adapter 1050 enables the computer system 1001 to connect and communicate with one or more networks 1060, such as a local area network (LAN), a wide area network (WAN), and/or a public network (e.g., the Internet) … Network adapter 1050 supports various communication protocols to ensure compatibility with network standards …”; para 222, “… Network 1060 is any computer network that can receive and/or transmit data. Network 1060 can include a WAN, LAN, private cloud, or public Internet, capable of communicating computer data over non-local distances by any technology for communicating computer data now known or to be developed in the future. Any connection depicted can be wired and/or wireless and may traverse other components that are not shown. In some embodiments, a network 1060 may be replaced and/or supplemented by LANs designed to communicate data between devices located in a local area, such as a Wi-Fi network. The network 1060 typically includes computer hardware such as copper transmission cables, optical transmission fibers, wireless transmission, routers, firewalls, switches, gateway computers, edge servers, and network infrastructure known now or to be developed in the future. Computer system 1001 connects to network 1060 via network adapter 1050 and bus 1030 …”), using one or more processors (para 98, “… The processor executes one or more LLMs on the vectorized data …”; para 99, “… The apparatus features a processor that receives real-time interaction content from the ongoing communication session between the source device (e.g., the customer's device) and the service provider device. The processor converts the received real-time interaction content into a vector representation, facilitating effective processing and analysis. The processor executes one or more LLMs on the vectorized real-time interaction content …”) …
receiving, by the one or more processors (para 98, “… The processor executes one or more LLMs on the vectorized data …”; para 99, “… The apparatus features a processor that receives real-time interaction content from the ongoing communication session between the source device (e.g., the customer's device) and the service provider device. The processor converts the received real-time interaction content into a vector representation, facilitating effective processing and analysis. The processor executes one or more LLMs on the vectorized real-time interaction content …), input from the user of a user device, wherein the input includes request information or a query associated with the conversation (FIG. 2D, items 212D, 214D; para 58, “… FIG. 2D, interaction content 214D is received from a communication session 212D between a source 206D of a source device 208D and a service provider device 210D …”; FIG. 3C, items 330, 332, 334, 336; para 81, “… The AI/ML production system 330 provides an application programming interface (API) 334, executed by an AI/ML server process 336 through which requests can be made … a request may include an AI/ML model 332 identifier to be executed … the AI/ML model 332 to be executed is implicit based on the type of request … a data payload (e.g., to be input to the model during execution) is included in the request …”; FIG. 9, item 910G, para 206, “… In 910G, the method may include receiving a search query from a software application, identifying at least one vectorized data within the vector database that correspond to the search query based on labels of the at least one vectorized data, and transmitting the at least one vectorized data to the software application …”);
extracting and processing, by the one or processors (para 98), the request information from the input to generate state information, wherein the state information includes intent information, emotional state information, slot information, and constraint information (para 103, “The system leverages LLMs to extract contextual attributes from the message content, including linguistic patterns, sentiment analysis, key terms, and conversational context. These attributes help understand the client's investment objectives, risk appetite, and market perceptions expressed during the con-versation. Based on the identified contextual attributes, the platform generates personalized investment recommendations and risk management strategies tailored to the client's needs and the prevailing market conditions. These responses may include suggestions for asset allocation, portfolio diversification, investment products, and potential risks associated with different strategies. The responses are then communicated to the client through various devices …”; para 107, “… the system analyzes communication patterns and behavioral cues within financial markets. It utilizes advanced linguistic models to analyze and interpret textual data, extracting valuable insights from various sources, including news articles, social media posts, and financial reports. The system extracts key information related to financial markets, companies, and economic indicators in real time. Utilizing LLMs identifies relevant keywords, sentiment indicators, and behavioral cues. The tool analyzes communication patterns and sentiment indicators and provides comprehensive insights into market sentiment. It identifies prevailing attitudes, emotions, and perceptions among investors, which can influence market dynamics and asset prices. It tracks investor sentiment by analyzing social media posts, forums, and other online discussions. It identifies emerging trends, hot topics, and investor sentiment shifts, allowing users to gauge market sentiment in real time …”);
routing, by the one or more processors (para 98), the input and the generated state information by enabling the virtual agent to facilitate the conversation between the user and the online system (para 42, “the host platform may use one or more artificial intelligence models to identify context of a conversation between a user and a service provider, such as a contact center of the service provider, a chatbot of the service provider, or the like … the context may include a mood of the user, a sentiment of the user, a tone of voice of the user, an item of interest, and the like. The system may use the contextual attributes to generate more accurate and customized responses which can be output during a live conversation between the user and the service provider …”; para 118, “The conversion is facilitated by executing an additional LLM, which transforms the textual interaction content and its associated contextual attributes into a numerical representation, effectively encoding the information into a vector format. The vector serves as a condensed representation of the conversation session, capturing both the content exchanged and the contextual nuances identified by the LLMs. Once the vector is generated, the processor labels the vector with identifiers corresponding to the plurality of contextual attributes identified earlier. These identifiers categorize and organize the vector data within a vector database, facilitating efficient storage and retrieval of conversational data … the vector is stored within the vector database along with a timestamp, providing temporal information about when the communication session occurred …”);
retrieving, by the one or more processors (para 98), information associated with the conversation from a Retrieval-Augmented Generation (RAG) database (FIG. 6A-6C, items 644, 660; para 28, “… FIGS. 6A-6C are diagrams illustrating a process of retrieval augmented generation (RAG) architecture for enhancing conversational responses generated by a LLM …”; para 43, “… The vectorized representation of the conversation may be managed within a vector database (or other storage) that is included with the LLM framework … the LLM framework may employ a retrieval augmented generation (RAG) framework for improving the efficiency of the outputs of the LLM models within the LLM framework …”), …
D’Agostino discloses a parallelized attention head architecture to generate a conversational mood. It would have been obvious to one of ordinary skill in the art before the effective filing date of the invention to include a parallelized attention head architecture to generate a conversational mood, as in D’Agostino; and to include customer assistance using AI-generated smart indicators, as in Mico, to improve and/or enhance the technology of an advice generation system, as in Ben David, because it would amount to combining elements that in the combination would perform the same function as they functioned separately. One of ordinary skill in the art before the effective filing date of the invention would have been motivated to combine the references to provide an AI framework identifying contextual attributes within a conversation between a customer, via a customer device, and a virtual agent or chatbot or contact center; a mood of the customer with respect to the item of interest; and specific concerns noted by the customer, dates, times, and the like, to output enhanced content about the item of interest to the customer during an active communication session, and improving the efficiency of the outputs of the large language models (LLM).
Ben David, Mico, and D’Agostino do not specifically disclose, however, Bathwal discloses providing, by the one or more processors (FIG. 16, items 1606, 1608, 1610; C/L 46/49-47/3, “… The machine 1600 includes processors 1606, … the processors 1610 (e.g., a central processing unit (CPU), a reduced instruction set computing (RISC) processor, a complex instruction set computing (CISC) processor, a graphics processing unit (GPU), a digital signal processor (DSP), an application-specific integrated circuit (ASIC), tensor processing unit (TPU), a radio-frequency integrated circuit (RFIC), another processor, or any suitable combination thereof) may include … a processor 1608 and a processor 1610 that may execute the instructions 1611. The term "processor" is intended to include multi-core processors 1606 that may comprise two or more independent processors (sometimes referred to as "cores") that may execute instructions 1611 contemporaneously. Although FIG. 16 shows multiple processors 1606, the machine 1600 may include a single processor with a single core, a single processor with multiple cores (e.g., a multi-core processor), multiple processors with a single core, multiple processors with multiple cores, or any combination thereof …”), via the virtual agent (C/L 6/27-47, “… When a user enters a search query into a search query input field (commonly referred to as a "search box," an "input box," or "search bar," the search engine processes the query by parsing and understanding the user's intent. The query processing component may include natural language processing (NLP) techniques to handle complex queries, synonyms, and context. Based on the processed query, the search engine retrieves the most relevant web pages from the index. The retrieval mechanism uses the ranking algorithm(s) to select and order the pages that best match the user's query. The search engine presents the retrieved results to the user through a user interface, typically a web page that displays a list of search results. Each result is simply a snippet of information about the web page, generally including a title, a URL, and a brief description or excerpt from the web page. The interface may also offer advanced search options and filters to refine the results. Modern search engines incorporate machine learning algorithms that analyze user interaction with search results (e.g., click-through rates, time spent on a page, etc.) to continuously improve the relevance and accuracy of the search results …”; C/L 24/47-25/3, “… the RL modeling method whose simplest setting treats a task as an agent interacting with the world in which the following assumptions hold: (1) the world is modeled as a set of states, (2) in each time step, the agent exists in a single state (S_t), (3) in each time step, the agent takes a single action (A_t), (4) in the next time step, the agent receives a scalar reward (R_ { t+ 1}) and finds itself in state (S_{t+l}), (5) the Markov property holds the distribution over (R_ {t+l }, S_ {t+l}) only depends on (S_t, A_t), and (6) the agent's behavior is episodic (it eventually ends in a terminal state (end of the episode) after a finite number of actions. Where "S" is "state," "A" is "action," "R" is "reward," and "t" is "time step." Each agent's behavior is dictated by a policy, such as a probability distribution over possible actions conditioned on the agent's current state. The goal of such an example is to learn a policy that maximizes the sum of rewards ("return") obtained by the agent … the sum can be (set gamma=1). According to examples using a reinforcement learning (RL) model for Natural Language Processing (NLP), text generation can be framed as an RL problem with the following characteristics: (1) the agent ("A") is the LLM, (2) the state (''S") is the prompt plus a sequence of tokens that has been generated at a certain point, …”), the outcome for the conversation to the user, wherein the LLM model is fine-tuned by generating the outcome using the state information and the user profile (FIG. 2, items 238, 242; C/L 14/56-67, “… an advanced interactive system utilizing generative large language models (LLMs) to fine-tune LLMs 242 into task-specific generative models … the indexing can be performed offline … this is performed directly in the browser and enables the system to provide real-time updated results as output 238 to the user 232. The system employs sophisticated algorithms to analyze the user's input and context, leveraging historical data, user preferences, and behavioral patterns to personalize subsequent interactions, as described and depicted in connection with the user information processing component 110 of FIG. 1 …”; FIG. 4, item 402; C/L 17/57-18/22, “ … FIG. 4 is an example block diagram 400 illustrating an example fine-tuning system 402 to fine-tune a pre-trained model (e.g., LLM) that is further trained on a smaller, task-specific dataset. This process allows the model to adapt its knowledge to the particularities of the desired task, … the model training and fine--tuning system architecture disclosed herein include one or more fine-tuning processes and/or use of proprietary datasets as components in the development and deployment of machine learning models, especially for tasks (e.g., search queries, summarization, etc.) that require a high degree of specificity and accuracy. The fine-tuning system 402 includes a machine learning technique where a pre-trained model is further trained (e.g., fine-tuned) on a new dataset that is generally smaller and/or more domain-specific than the data used in the initial model training. The fine-tuning system 402 allows the model to adapt to nuances, specific requirements, or the like of the new task or domain. The block diagram 400 illustrates components of the fine-tuning system 402, including a base model selector 408, proprietary dataset creator 410, data preprocessor 412, model adaptor 414, trainer and validator 416, evaluator 418, and deployment component 420. The base model selector 408 of the fine-tuning system 402 begins with the selection of a large, pre-trained model that has been trained on a vast and diverse dataset. This model serves as the starting point and has already learned a rich representation of language features … the fine-tuning system 402 would choose an LLM that has been pre-trained on a large and diverse corpus of data …”).
Bathwal discloses enhanced searching using fine-tuned machine learning models. It would have been obvious to one of ordinary skill in the art before the effective filing date of the invention to include enhanced searching using fine-tuned machine learning models, as in Bathwal; to include a parallelized attention head architecture to generate a conversational mood, as D’Agostino; and to include customer assistance using AI-generated smart indicators, as in Mico, to improve and/or enhance the technology of an advice generation system, as in Ben David, because it would amount to combining elements that in the combination would perform the same function as they functioned separately. One of ordinary skill in the art before the effective filing date of the invention would have been motivated to combine the references to provide special purpose machines that use large language models and generative artificial intelligence for summarization, more specifically, to provide enhance search capabilities using fine-tuned machine learning models, and by utilizing natural language processing techniques enabling the extraction of meaningful patterns, sentiment, and entities from text, which can improve the accuracy and contextuality of search results.
Regarding claims 2 and 17, Ben David, Mico, D’Agostino, and Bathwal disclose the limitations of claims 1 and 16. Ben David, D’Agostino, and Bathwal do not specifically disclose, however, Mico discloses the method of claim 1, further comprising:
analyzing the response to the alternative proposition from the user to generate a future proposition (para 33, “… an artificial intelligence based configurator of the disclosed system may set up the smart indicators based on certain scenario (e.g., festival traffic, etc.) … the artificial intelligence based configurator may have a business rule that triggers a particular indicator set to appear and/or the order in which they appear. The artificial intelligence based configurator may dynamically change the order of the smart indicator(s) based on the response rate and/or the historical knowledge of the customer…”); and
implementing the future proposition in a future conversation associated with the user or another user (para 33, “… when the neural network activates (e.g., during festivals, sports tournaments, etc.) for a particular indicator that has the same kind of user journey, the artificial intelligence algorithm may dynamically reorder future expressions…”; para 53, “… the response generation module 206 may concurrently update the query index 204 each time the user intent 110 and/or derived context information 112 do not map with existing query index 204 to include the newly identified user intent 110 and/or context information 112. Subsequently, the response generation module 206 may present with the newly generated smart indicator 208 in real time when a similar context and/or situation is identified in the future …”).
Mico discloses customer assistance using AI-generated smart indicators. It would have been obvious to one of ordinary skill in the art before the effective filing date of the invention to include customer assistance using AI-generated smart indicators, as in Mico; to include enhanced searching using fine-tuned machine learning models, as in Bathwal; and to include a parallelized attention head architecture to generate a conversational mood, as D’Agostino, to improve and/or enhance the technology of an advice generation system, as in Ben David, because it would amount to combining elements that in the combination would perform the same function as they functioned separately. One of ordinary skill in the art before the effective filing date of the invention would have been motivated to combine the references to provide an AI based configurator integrated within the conversational software application, including an intelligent indicator generation algorithm, that automatically changes the selection and/or the order of button-based options within the chatbot, and providing recommendations to predict the most likely indicator options that will lead to satisfactory response to a customer’s concerns and a successful outcome for the customer, as well as any future outcome.
Regarding claims 4 and 12, Ben David, Mico, D’Agostino, and Bathwal disclose the limitations of claims 1 and 11. Ben David further discloses the method of claim 1, wherein the alternative engagement strategy based on consultation includes one or more alternative engagement strategies derived by prioritizing one or more of emotional state information, constraint information, and intent information associated with the user (C/L 3/21-29, “… Relevant data is gathered from the user, including their current financial state, financial reports, and necessary trans-actional data. This data is input into the LLM and forms the basis for the personalized advice. Output from the LLM is analyzed by a classification model to characterize the user's 25 intent within a known ontology, enabling the system to address the specific problems raised by the user and offer targeted advice that is tailored to the customer's unique circumstances and requirements …”; FIG. 3, item 310; C/L 6/44-47, “… At Step 310, an intent generated by a large language 45 model is received as a first input to the advice planner. The Intent is generated by the large language model from a text received from a user device …”).
Regarding claim 5, Ben David, Mico, D’Agostino, and Bathwal disclose the limitations of claim 1. Ben David, D’Agostino, and Bathwal do not specifically disclose, however, Mico discloses the method of claim 1, wherein the alternative engagement strategy is based on one or more of a credibility appeal, a logical appeal, and an emotional appeal (para 54, “… The database 122 of the automated chatbot server 102 may automatically maintain semantic data by structuring the interactional data 118 and context information 112 derived from the user intent 110 in order to represent it in a specific logical way each time a user interacts through the dynamic chatbot 104. The semantic data may help identify the set of conditions that will trigger the artificial intelligence based configurator 126 to automatically generate and/or display a relevant set of indicators (e.g., smart indicator 208) for the user 210 …”; ).
Mico discloses customer assistance using AI-generated smart indicators. It would have been obvious to one of ordinary skill in the art before the effective filing date of the invention to include customer assistance using AI-generated smart indicators, as in Mico; to include enhanced searching using fine-tuned machine learning models, as in Bathwal; and to include a parallelized attention head architecture to generate a conversational mood, as D’Agostino, to improve and/or enhance the technology of an advice generation system, as in Ben David, because it would amount to combining elements that in the combination would perform the same function as they functioned separately. One of ordinary skill in the art before the effective filing date of the invention would have been motivated to combine the references to provide the hardware and software for customer assistance using artificial intelligence (AI)-generated smart indicators, which may predict the most likely context-based indicator options that will lead to a successful outcome for the customer, and automatically maintaining semantic data by structuring the interactional data and context information, in order to present it in a specific logic way each time a user interacts through the dynamic chatbot.
Regarding claims 6 and 20, Ben David, Mico, D’Agostino, and Bathwal disclose the limitations of claims 1, 16, and 19. Ben David further discloses the method of claim 1, wherein the additional information includes one or more of nonverbal cues, situation information, and emotional state information associated with the user transaction (FIG. 3, item 310; C/L 6/44-60, “… At Step 310, an intent generated by a large language model is received as a first input to the advice planner. The Intent is generated by the large language model from a text received from a user device. The text data, such as in documents, emails, social media posts, and web pages, can be provided to the LLM through an interface, either directly through a user interface of the user device, or through an API call from the advice planner. Unlike structured data, text data lacks a fixed format and can vary in length, style, and content. It is represented as sequences of characters encoded in a specific scheme. Text data can be represented as a series of characters encoded in a specific character encoding scheme, such as ASCII, UTF-8, or Unicode. The text input serves as the foundation for generating the intent by the LLM. Based on the text data, the LLM generates an intent that captures the user's purpose or goal …”).
Regarding claim 7, Ben David, Mico, D’Agostino, and Bathwal disclose the limitations of claim 1. Ben David, D’Agostino, and Bathwal do not specifically disclose, however, Mico discloses the method of claim 1, wherein the alternative proposition includes one or more of targeted advice associated with the conversation, a counteroffer, a discounted price, an alternative service, an alternative product, an additional service, an additional product, and a modified service, or a modified product (para 35, “… The buyers and sellers may communicate using the disclosed framework in an optimized way that allows various aspects of decisions (e.g., purchasing decisions) to be negotiated and adjusted nearly instantly … the consumers use their client devices (e.g., laptop, smart phone, tablet, etc.) to anonymously organize and set up automated, on-going recommendation plans executed by multi-criteria decision-making negotiation agents on a server, referred to herein as "buyer AI negotiators" ("AI buyers," "AI buyer agent," "consumer bot," or "purchasing bot" for short), for purchasing particular products or services, either online or through physical kiosks or storefronts … some examples allow retailers to set up automated, on-going recommendation terminals or selling campaigns executed by multi-criteria decision-making negotiation agents on a server, referred to herein as "seller AI negotiators" (or "AI sellers" "AI seller agent" for short), that offer the seller's products or services and automatically recommendation with the buyer AI negotiators. The various embodiments may operate by combining large data sets of customer and consumer data from internal and/or external sources within an organization and applying intelligent, iterative processing algorithms to learn from patterns and features in the data that they analyze …”).
Mico discloses customer assistance using AI-generated smart indicators. It would have been obvious to one of ordinary skill in the art before the effective filing date of the invention to include customer assistance using AI-generated smart indicators, as in Mico; to include enhanced searching using fine-tuned machine learning models, as in Bathwal; and to include a parallelized attention head architecture to generate a conversational mood, as D’Agostino, to improve and/or enhance the technology of an advice generation system, as in Ben David, because it would amount to combining elements that in the combination would perform the same function as they functioned separately. One of ordinary skill in the art before the effective filing date of the invention would have been motivated to combine the references to provide the hardware and software for customer assistance using artificial intelligence (AI)-generated smart indicators, which may predict the most likely context-based indicator options that will lead to a successful outcome for the customer.
Regarding claim 8, Ben David, Mico, D’Agostino, and Bathwal disclose the limitations of claim 1. Ben David, D’Agostino, and Bathwal do not specifically disclose, however, Mico discloses the method of claim 1, wherein the alternative proposition includes one or more of a persuasion, a nudge, a negotiation, and advice associated with the conversation (para 34, “… a server-implemented framework is disclosed which automates the discovery and negotiation of product and service sales online and offline based on buyer- and seller-defined parameters and elasticity thresholds. Artificial intelligence (AI) negotiation agents operate on behalf of the buyers and sellers to recommend potential options, automatically and anonymously negotiate towards the best satisfaction and outcome for their respective users based on the parameters set by the users to be important and also based on market conditions. The AI negotiation agents join a multi-stage negotiation session until sufficiently improved recommendations are obtained for particular products and/or services. These negotiated, improved, offers are then transmitted to the buyers and sellers for acceptance. The AI agent for sellers optimizes sales strategy and effectiveness while the AI agent for buyers improves purchasing decision-making to enable better outcomes (financial, customer satisfaction, experience, etc.) with minimal effort…”).
Mico discloses customer assistance using AI-generated smart indicators. It would have been obvious to one of ordinary skill in the art before the effective filing date of the invention to include customer assistance using AI-generated smart indicators, as in Mico; to include enhanced searching using fine-tuned machine learning models, as in Bathwal; and to include a parallelized attention head architecture to generate a conversational mood, as D’Agostino, to improve and/or enhance the technology of an advice generation system, as in Ben David, because it would amount to combining elements that in the combination would perform the same function as they functioned separately. One of ordinary skill in the art before the effective filing date of the invention would have been motivated to combine the references to provide the hardware and software for customer assistance using artificial intelligence (AI)-generated smart indicators, which may predict the most likely context-based indicator options that will lead to a successful outcome for the customer.
Regarding claims 9 and 14, Ben David, Mico, D’Agostino, and Bathwal disclose the limitations of claims 1 and 11. Ben David further discloses the method of claim 1, wherein generating the initial proposition communication includes providing the initial engagement strategy to the LLM model to generate the initial proposition communication (FIG. 3, item 350; C/L 8/23-37 “… At Step 350, the plan is forwarded to the large language model (LLM). The large language model receives the plan as input and generates advice in a natural language format as output. The plan generated by the advice planner is forwarded to the LLM as input. This can be achieved through suitable protocols or APIs that facilitate the transmission of data between the advice planner and the LLM. The plan can be formatted according to a prompt template, influencing the LLM. to interpret inputs in an equivalent manner, and produce similarly formatted output. Upon receiving the plan as input, the LLM employs its natural language processing capabilities to analyze and understand the content of the plan, and to generate advice. The advice is generated in a natural language format …”).
Regarding claims 10 and 15, Ben David, Mico, D’Agostino, and Bathwal disclose the limitations of claims 1 and 11. Ben David further discloses the method of claim 1, wherein generating the alternative proposition communication includes providing the alternative engagement strategy to the LLM model to generate the alternative proposition communication (FIG. 4, items 410, 420, 430; C/L 8/54-55 “… At Step 410, a prompt to the LLM is generated from a template and the intent …”; C/L 8/65-9/1, “ … Upon receiving the prompt as input, the LLM processes the prompt and generates a second set of action logic that corresponds to the intent and aligns with the structure of the prompt template …”; C/L 9/4-8, “… At Step 420, the advice library is updated to include the second set of action logic. The advice library is updated to include the newly generated second set of action logic, ensures that the advice library remains comprehensive and up to date with the latest information and recommendations …”; C/L 9/15-25, “ … At Step 430, the advice planner is retrained from both the first set of action logic and the second set of action logic. The advice planner incorporates both the first set of action logic, obtained from previous steps, and the newly generated second set of action logic. The retraining process involves updating the advice planner's algorithms, models, or parameters using the combined knowledge from the first and second sets of action logic. The retraining ensures that the advice planner can account for the relevant information unearthed from the LLM to generate accurate and effective advice …”).
Regarding claim 19, Ben David, Mico, D’Agostino, and Bathwal disclose the limitations of claim 16. Ben David further discloses the system of claim 18, wherein the alternative engagement strategy based on consultation includes one or more alternative engagement strategies derived by prioritizing one or more of emotional state information, constraint information, and intent information associated with the user (C/L 4/25-33, “… account(s) (110) may include profile information such as the user's personal information, contact details, preferences, as well as any other demographic or identification data associated with their account, in the context of financial advice or planning, the account may financial data, such as user's financial records, transaction history, income, expenses, assets, account balances, liabilities, historical data, usage patterns, and/or other financial metrics…”). For examination purposes, claim 19 is being interpreted as depending from claim 16.
Regarding claim 21-23, Ben David, Mico, D’Agostino, and Bathwal disclose the limitations of claims 1, 11, and 16. Ben David, Mico, and D’Agostino do not specifically disclose, however, Bathwal discloses the method of claim 1, further comprising:
retrieving a dataset from an internal database (FIG. 6A, items 642, 644, 648; para 132, “… FIG. 6A illustrates a process 600A of a retriever 642 within a RAG-based architecture which can identify a subset of vectors within a vector storage 644 and augment an input to an LLM 646 with the subset 648 of vectors …”; FIG. 9C, item 906C; para 200, “… In method 906C, the method may include retrieving a subset of vectors from a plurality of vectors stored in a vector database based on the search criteria of the interaction content, wherein the subset of vectors includes previous interaction content with the service provider … );
performing data sampling on the retrieved dataset (para 48, “… the interaction content may be analyzed for contextual attributes. The one or more LLMs 122 may annotate the interaction content with the contextual attributes prior to converting the interaction content into a vector …”; FIG. 3B, items 332, 344; para 75, “… Features of the data are identified and extracted 346 … a feature of the data is internal to the prepared data from step 344 … a feature of the data requires a piece of prepared data from step 344 to be enriched by data from another data source to be useful in developing an AI/ML model 332 … identifying features is a manual process or an automated process using at least one of the elements, functions described or depicted herein. Once the features have been identified, the values of the features are collected into a dataset that will be used to develop the AI/ML model 332 …);
generating an intent by processing the sampled dataset along with the user profile of the user (para 107, “… the system analyzes communication patterns and behavioral cues within financial markets. It utilizes advanced linguistic models to analyze and interpret textual data, extracting valuable insights from various sources, including news articles, social media posts, and financial reports. The system extracts key information related to financial markets, companies, and economic indicators in real time. Utilizing LLMs identifies relevant key-words, sentiment indicators, and behavioral cues. The tool analyzes communication patterns and sentiment indicators and provides comprehensive insights into market sentiment. It identifies prevailing attitudes, emotions, and perceptions among investors, which can influence market dynamics and asset prices. It tracks investor sentiment by analyzing social media posts, forums, and other online discussions. It identifies emerging trends, hot topics, and investor sentiment shifts, allowing users to gauge market sentiment in real time …”; para 150, “… The search retrieves a subset of vectors that closely match the user's unique context and requirements. These vectors serve as valuable insights and references for generating personalized investment recommendations. The retrieved subset of vectors is then fed into LLMs, which analyze the historical data and extract relevant patterns, trends, and insights related to investment strategies, asset allocations, risk management techniques, and market dynamics. Leveraging the insights from the LLM analysis, the system generates personalized investment recommendations tailored to the user's financial situation and objectives. These recommendations encompass asset allocation suggestions, portfolio diversification strategies, investment product recommendations, risk mitigation techniques, and long-term financial planning advice. The personalized investment recommendations are presented to the user through the chosen communication channel, be it a web interface, mobile app notification, email, or personalized report. The service offers additional features, including interactive visualization tools, scenario analysis, investment performance tracking, and periodic portfolio reviews to enhance user engagement and satisfaction … );
performing an intent consistency check by matching the generated intent with a subsequent intent (para 41, “… the system may store the vector with the labels in a database, such as a vector database. By labelling the vectors with the contextual attributes, a search process may be used to retrieve vectors that match a search criteria, such as a specific value for a contextual attribute. The vectors can be input to an AI model (such as an AI conversational model), a LLM, a chatbot, or the like, which can generate custom instructions, responses, verifications, and the like, to output during a live/real-time (or near-real time) communication session …”; para 123, “… an apparatus is configured to receive a search query from a software application, identify one or more vectors within the vector database that correspond to the search query based on labels of the vectors, and transmit the one or more vectors to the software application. The functionality enables the apparatus to retrieve relevant conversation data from the vector database in response to search queries initiated by a software application. The software application sends a search query to the processor, specifying criteria or keywords for retrieving relevant conversation data. Upon receiving the search query, the processor utilizes the labels associated with vectors stored in the vector database to identify one or more vectors that match the search criteria by comparing the search query with the labels of stored vectors to determine relevance. Once the relevant vectors are identified, the processor transmits the vectors to the software application, enabling the application to access and utilize the corresponding conversation data. The data transmission can send the vectors directly to the software application or provide access to the vectors through an application programming interface (API) … ); and
generating a dialogue flow sequence based on the generated intent and the matched subsequent intent (para 162, “… an apparatus comprising a processor and a memory receives interaction content from a communication session between a source device and a service provider device. The interaction content includes text-based dialogue between a user and a service provider, such as a customer support representative or a chatbot. The processor then executes a LLM on the interaction content. The LLM is equipped with a plurality of attention heads configured to simultaneously identify a mood and an item of interest from the interaction content. The attention heads function as specialized modules within the neural network, each focusing on different aspects of the conversation. As the conversation unfolds, the processor feeds the interaction content into the LLM, and each attention head independently analyzes distinct contextual attributes, such as mood and item of interest. Once the analysis is complete, the processor generates a response to the interaction content based on the identified mood and item of interest. The response can include personalized recommendations, assistance, and information tailored to the user's mood and preferences. The processor outputs the response to at least one of the source devices and the service provider device during the ongoing communication session, allowing for seamless integration of the generated response into the conversation flow between the user and the service provider. Additionally, the apparatus can adapt to the evolving nature of the conversation by continuously analyzing and processing new interaction content as it becomes available during the communication session … ).
Bathwal discloses enhanced searching using fine-tuned machine learning models. It would have been obvious to one of ordinary skill in the art before the effective filing date of the invention to include enhanced searching using fine-tuned machine learning models, as in Bathwal; to include a parallelized attention head architecture to generate a conversational mood, as D’Agostino; and to include customer assistance using AI-generated smart indicators, as in Mico, to improve and/or enhance the technology of an advice generation system, as in Ben David, because it would amount to combining elements that in the combination would perform the same function as they functioned separately. One of ordinary skill in the art before the effective filing date of the invention would have been motivated to combine the references to provide special purpose machines that use large language models and generative artificial intelligence for summarization, more specifically, to provide enhance search capabilities using fine-tuned machine learning models, and by utilizing natural language processing techniques enabling the extraction of meaningful patterns, sentiment, and entities from text, which can improve the accuracy and contextuality of search results.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure:
Wu et al (U. S. Patent Application Publication No. 20210150545 A1) – Providing Product Recommendation In Automated Chatting
Wu discloses method and apparatus for facilitating product recommendation in automated chatting. In some implementations, it may be determined that a terminal device is within a predefined area, a user identity may be obtained through communicating with a chatbot on the terminal device, product recommendation information associated with the user identity may be determined and provided to the chatbot. In some implementations, a first message may be received in a chat flow, a response to the first message may be provided for indicating at least one product determined based at least on the first message, a second message including a comment on the at least one product may be received, and a user preference on the at least one product may be determined based at least on the second message.
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 STEVEN CHISM whose telephone number is (571) 272-5915. The examiner can normally be reached during 9:00 AM – 3:00 PM Monday – Thursday, EST.
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, Ryan D. Donlon can be reached (571) 270-3602. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of an application may be obtained from the Patent Application Information Retrieval (PAIR) system. Status information for published applications may be obtained from either Private PAIR or Public PAIR. Status information for unpublished applications is available through Private PAIR only. For more information about the PAIR system, see https://ppair-my.uspto.gov/pair/PrivatePair. Should you have questions on access to the Private PAIR system, contact the Electronic Business Center (EBC) at 866-217-9197 (toll free). If you would like assistance from a USPTO Customer Service Representative or access to the automated information system, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/STEVEN R CHISM/Examiner, Art Unit 3692
/RYAN D DONLON/Supervisory Patent Examiner, Art Unit 3692 August 7, 2026
Prosecution Insights