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 .
Claim Rejections - 35 USC § 112
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 3, 8, 9, 16, 17, 18 are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-
AIA 35 U.S.C. 112, the applicant), regards as the invention.
Regarding claim 3 and 18, the claims recites “each of the series of prompts includes each prompt from the group of prompts and a second indication, wherein the second indication indicates to adjust each prompt based on each IDR type from the group of IDR types” at lines 6-7 and line 8 for claim 18. This is ambiguous because the relationship between prompts and IDR types is not clearly outlined. It is unclear whether each prompt is adjusted based on every IDR type, or whether there is a one-to-one correspondence between a respective prompt and a respective IDR type. As such, the claim as drafted makes it unclear about the relationships between prompt and IDR types and therefore is rejected under 112(b) as being indefinite.
Regarding the claim 8, 9 , 16, and 17 the claims recites ”previously generated IDR’s”. It is unclear what the claim is implying because the specification treats IDR’s as received requests/documents rather than things that are “generated”. The specification further talks about “generating IDR responses” based on previously stored IDR responses rather than generating “previous IDR’s”. Therefore, claim 8, 9, 16 and 17 are unclear and are rejected under 112(b) as being indefinite.
Appropriate correction is required.
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claim [ 1, 4, 5 ] are rejected under 35 U.S.C. 103 as being unpatentable over Li (US12217318, hereinafter Li) in view of Gray (US11769017B1, hereinafter Gray).
Regarding claim 1, Li teaches
A device, comprising: a processing system including a processor; and a memory that stores executable instructions that, when executed by the processing system, facilitate performance of operations, the operations comprising:
[Column 2, lines 57-60 "In some embodiments, a first non-transitory computer-readable storage medium is provided, the first non-transitory computer-readable storage medium storing instructions for determining the composition of document bundles, the instructions configured to be executed by one or more processors of a system to cause the system to: receive first input data comprising a document bundle;"];
[Column 26, lines 26- 30 "Storage 2640 can be any suitable device that provides storage, such as an electrical, magnetic, or optical memory, including a random access memory (RAM), cache, hard drive, CD-ROM drive, tape drive, or removable storage disk."]
receiving, over a communication network, an information document request (IDR) from a first computing device associated with a first user;
[Column 9, lines 37-40 "As described herein, a system for providing AI-augmented auditing platforms may be configured to receive one or more documents as input data and to perform automated processing of the input documents" where one or more documents correspond to Information document request];
[Column 10, lines 55-59 “Context data may be received by the system from any one or more suitable data sources, may be indicated by one or more user inputs detected by the system, and/or may be deduced by one or more data processing models of the system”].
determining a first IDR type associated with the IDR resulting in a first determination;
[Column 9, lines 60- 63 "In some embodiments, a system for automated document processing may be configured to perform automated document classification (e.g., classifying documents according to different document types) and/or document bundling" where determining the document type(IDR type) associated with the received document is the first determination].
Providing an IDR
[Column 9, lines 37-40 "As described herein, a system for providing AI-augmented auditing platforms may be configured to receive one or more documents as input data and to perform automated processing of the input documents" where one or more documents correspond to Information document request].cl
However, Li fails to expressly recite selecting a first adjusted prompt from a group of adjusted prompts based on the first determination; selecting a first artificial intelligence (AI) model from a group of AI models based on the first determination resulting in a first AI model selection; providing the IDR, the first adjusted prompt, and the first AI model selection to an AI software application; generating a response associated with the IDR utilizing the AI software application based on the IDR, the first adjusted prompt, and the first AI model selection; and providing, over the communication network, the response associated with the IDR to the first computing device associated with the first user.
But Gray teaches selecting a first adjusted prompt from a group of adjusted prompts based on the first determination;
[Column 5, lines 1-17 "For example, if it is determined that the user is familiar with “X content”, a prompt of “assuming the user is familiar with [description of the certain content] answer [query]” can be processed using the LLM in generating the NL based summary. In contrast, if it is determined that the user is not familiar with any content responsive to the query, a prompt of “answer [query]” can be processed using the LLM in generating the NL based summary. As another example, if it is determined that the user is familiar with “X content”, a prompt of “assuming the user is familiar with [description of the certain content] summarize [search result document content]” can be processed using the LLM in generating the NL based summary. In contrast, if it is determined that the user is not familiar with any content responsive to the query, a prompt of “summarize [search result document content]” can be processed using the LLM in generating the NL based summary” where determining condition corresponds to determination and selecting one prompt vs another implies selecting a first adjusted prompt.];
[Column 23, lines 38-46 "At sub-block 460A, the system generates the revised NL based summary using the same LLM as used in block 454, but using a revised input with a revised prompt that reflects familiarity with the content of the SRD(s) interacted with. For example, the prompt used in block 454 could be “create a summary of the following”, whereas the prompt used in sub-block 460A could be “create a summary of the following and assuming the user already knows X”, where “X” is a description of the content of the SRD(s) interacted with. In some implementations, the system can generate the description “X” to conform to some or all of the content of the SRD(s) interacted with"].
selecting a first artificial intelligence (AI) model from a group of AI models based on the first determination resulting in a first AI model selection;
[Column 9, 10, lines 64-67, lines 1-3 "The LLM selection engine 132 can, in response to receiving a query, determine which, if any, of multiple generative model(s) (LLM(s) and/or other generative model(s)) to utilize in generating response(s) to render responsive to the query" where multiple generative models correspond to the group of AI models, the determination of which generative model to utilize corresponds to selecting the first AI model, and the determined generative model corresponds to the first AI model selection.].
[Column 29, lines 51- 55 "Yet another example, a first prompt can be processed (along with additional content) using a first LLM and a second prompt can be processed (optionally along with the same additional content) using a second LLM" where the additional content is a third component, the first prompt corresponds to the first adjusted prompt, and the first LLM corresponds to the first AI model selection];
[Column 37, lines 44-46 "In some implementations, the candidate generative models include a given large language model (LLM) paired with a first prompt to be used in response generation…"];
[Column 9, lines 20-25 "Further, the client device 110 and/or the NL based response system 120 can include one or more memories for storage of data and/or software applications, one or more processors for accessing data and executing the software applications, and/or other components that facilitate communication over one or more of the networks 199."];
generating a response associated with the IDR utilizing the AI software application
[Column 10, lines 8- "The LLM input engine 134 can, in response to receiving a query, generate LLM input that is to be processed using an LLM in generating an NL based response to the query. As described herein, such content can include query content that is based on the query and/or additional content" where additional content using the LLM to generate the NL based summary corresponds to the third component];
[Column 37, lines 44-46 "In some implementations, the candidate generative models include a given large language model (LLM) paired with a first prompt to be used in response generation…"];
[Column 23, lines 58- 62 "At sub-block 4606, the system generates the revised NL based summary using an additional LLM, relative to the one used in block 454, that is fine-tuned based on a prompt that reflects familiarity with content of the SRD(s) interacted with"];
[Column 10, lines 17-19 "The LLM response generation engine 136 can process LLM input, that is generated by the LLM input engine 134, using an LLM to generate an NL based summary"];
[Column 9, lines 20-25 "Further, the client device 110 and/or the NL based response system 120 can include one or more memories for storage of data and/or software applications, one or more processors for accessing data and executing the software applications, and/or other components that facilitate communication over one or more of the networks 199."].
providing, over the communication network, the response associated with the IDR to the first computing device associated with the first user.
[Column 32, lines 31-34 "The method further includes generating a natural language (NL) based summary using the LLM output, and causing the NL based summary to be rendered at the client device in response to the query" where the NL based summary is the response].
It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Li with Gray because Gray teaches that different generative models may perform better in different situations and enables use of the output from the model best suited for the current situation. This combination would allow Li’s classified audit documents to be processed using an appropriately selected AI model and prompt, thereby improving the accuracy and suitability of AI generated response for different IDR types and content.
Regarding claim 4, the rejection of claim 1 is incorporated. Li and Gray teach all limitations of the current invention as stated above.
Li further teaches the device of claim 1, wherein the operations comprise, in response to receiving the IDR, processing the IDR for the AI software application.
[Column 9, lines 37-40 "As described herein, a system for providing AI-augmented auditing platforms may be configured to receive one or more documents as input data and to perform automated processing of the input documents" where one or more documents correspond to Information document request];
[Column 10, lines 3-12 "The system may be configured to receive data representing one or more documents and to apply one or more AI methods to the received data in order to recognize and extract information from said documents and in order to classify and/or cluster said documents. The AI methods may be configured to perform analyses on the basis of substantive document content (e.g., characters, text, and/or images in the documents), on the basis of metadata stored as a part of or in association with said document, and/or on the basis of context data associated with said documents"].
Regarding claim 5, the rejection of claim 1 is incorporated. Li and Gray teach all limitations of the current invention as stated above.
Li further teaches the device of claim 4, wherein the processing of the IDR comprises
performing optical character recognition (OCR) on the IDR.
[Column 4, lines 33- 44 "In some embodiments, a third system is provided, the third system being for extracting information from documents, the third system comprising one or more processors configured to cause the third system to: receive a data set comprising a plurality of electronic documents; apply a set of data conversion processing steps to the plurality of electronic documents to generate a processed data set comprising structured data generated based on the plurality of electronic documents, wherein applying set of data conversion processing steps comprises applying one or more deep-learning-based optical character recognition (OCR) models;"].
Claim [ 2, 3, 6, 8, 9] are rejected under 35 U.S.C. 103 as being unpatentable over Li (US12217318, hereinafter Li) in view of Gray (US11769017B1, hereinafter Gray) and in further view of Scherle (US20250173550A1, hereinafter Scherle) and in further view of Beauchamp ( US20240256762A1, hereinafter Beauchamp) and in further view of Schillace (US12505296B2, hereinafter Schillace).
Regarding claim 2, Li further teaches
The device of claim 1, wherein the operations comprise, prior to receiving the IDR from the computing device associated with the first user:
identifying the first IDR type;
[Column 9, lines 60- 63 "In some embodiments, a system for automated document processing may be configured to perform automated document classification (e.g., classifying documents according to different document types) and/or document bundling" where determining the document type(IDR type) associated with the received document is the first determination].
However, Li in view of Gray fails to expressly recite providing the first IDR type to a second computing device associated with a second user;
But Scherle recites providing the first IDR type to a second computing device associated with a second user;
[0089 "At operation 318 of the method 300, the classification results are presented in a user interface using the user interface component 212 of the data classification system 128. For example, the user 108 may be presented with the classification results for review or validation in a user interface provided at the user device 106" where classification results indicates the type of request];
[0166 “The applications 720 may utilize built in operating system functions (e.g., kernel 728, services 730 or drivers 732), libraries (e.g., system libraries 734, API libraries 736, and other libraries 738), and frameworks/middleware layer 718 to create user interfaces to interact with users of the system” where user interfaces to interact with users indicates multiple users];
[0022 "The classification scheme may be used to guide or indicate to a machine learning model how to classify the process metadata. For example, the classification scheme may indicate that each attribute should be classified as either being a personal data attribute or not being a personal data attribute. Alternatively, or additionally, the classification scheme may indicate that each attribute should be classified into one of a set of predetermined categories (e.g., “text,” “numerical,” “categorical,” or “date”). The classification scheme may provide a standardized taxonomy to categorize items using consistent labels across different automated processes (e.g., within an organization). The instruction may thus include or reference the predetermined classification scheme"].
It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Li in view of Gray with Scherle’s technique of providing an IDR to a computing device associated with another user and receiving a user input in response, because doing so would enable the identified IDR and its associated type information to be communicated to an appropriate user for review and further action. The combination would advantageously facilitate collaborative processing of IDR’s by allowing relevant IDR information to be distributed to the appropriate user.
Li in view of Gray and in further view Scherle do not teach receiving first user-generated input from the second computing device associated with the second user, wherein the first user-generated input indicates a first prompt; providing a prompt to the AI software application, wherein the prompt includes the first prompt and a first indication, wherein the first indication indicates to adjust the first prompt based on the first IDR type; and generating the first adjusted prompt utilizing the AI software application.
However, Beauchamp teaches receiving first user-generated input from the second computing device associated with the second user, wherein the first user-generated input indicates a first prompt;
[0086 "In this example, each user element 622 is provided with a single-instruction submit option 624 and a single-instruction cancel option 626. Selection of the single-instruction submit option 624 may cause the instruction contained in the user element 622 to be provided to the prompt generator 500 together with the corresponding selected portion of text 614 … This may enable generation of a first prompt based on user actions"];
[0083 "If multiple users (using respective multiple user devices) are involved in composing and/or editing the text, each user may independently access the text-editor 550 (e.g., using multiple instances of the text-editor 550) and/or may independently access services provided by the text-editor 550, to independently compose and/or edit the text." where respective multiple devices indicates multiple users which corresponds to a second user];
[0015 "In an example of any of the preceding example systems, the processing unit may be configured to execute computer-readable instructions to further cause the system to: generate the prompt including at least the portion of the annotated block of text, the prompt also including an instruction to cause the LLM to further annotate the annotated block of text in accordance with at least one inserted text-editing instruction…."]
It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Li in view of Gray in view of Scherle with the teachings of Beauchamp because doing so would allow the user receiving the IDR information to provide a prompt that is subsequently adapted in view of the information associated with the IDR type. This combination would advantageously provide prompts tailored to the particular type of IDR being processed, therefore would improve the relevancy and effectiveness of subsequent AI processing.
Li in view of Gray and in view Scherle and in further view of Beauchamp do not teach providing a prompt to the AI software application, wherein the prompt includes the first prompt and a first indication, wherein the first indication indicates to adjust the first prompt based on the first IDR type; and generating the first adjusted prompt utilizing the AI software application.
However, Schillace teaches providing a prompt to the AI software application, wherein the prompt includes the first prompt and a first indication, wherein the first indication indicates to adjust the first prompt based on the first IDR type;
[Column 5, lines 12- 20 "The prompt generator 128, receives the task objective and task request and utilizes them to generate one or more prompts for the ML model. The prompt generator 128 may generate one or more prompts that when processed a ML model, such as a generative large language model (LLM), provide sufficient context for the ML model to generate model output responsive to the task objective associated with the input" where the task request can correspond to the first prompt, the task objective reflecting IDR type(context associated with the task objective) corresponds to the first indication because the task objective specifies the context to be incorporated into the prompt, the context represented by the task objective is based on the first IDR type identified by Li];
[Column 5, liens 24- 30 "Thus, the one or more prompts encompass the semantic context of the task objective and task request so that the ML model can generate model output responsive to the requested task and/or intent without requiring additional training or fine-tuning of the model prior to generating model output responsive to the task or intent"];
[Column 9, lines 25- 31 "At operation 204, a task objective may be determined, for example, by a task objective module (e.g., task objective module 126) based on the input and task request. The task objective may encapsulate the general intent or specific meaning of the input and may be utilized to assist in generating one or more prompts specifically related to that objective" where task objective is determined based on the input and task request and input could be the document request from Li's reference];
[Column 5, lines 55-67 "The prompts may be generated in a variety of ways. In one example, an application and/or ML model from model repository 130 may analyze the task objective and input to select from one or more prompt templates stored in data store 106 with which to populate the prompt. In an alternative example, NL processing tools may be utilized to analyze the input and task objective to determine one or more associated prompt templates and/or populate a prompt. In a further example, the prompt generator 128 may use the ML model directly to either write a prompt, rewrite a prompt, and/or expand a seed prompt to determine one or more associated prompt templates and/or populate a prompt"]
generating the first adjusted prompt utilizing the AI software application.
[Column 9, lines 63-67 "In a further example, the prompt generator 128 may use the ML model directly to either write a prompt, rewrite a prompt, and/or expand a seed prompt to determine one or more associated prompt templates and/or populate a prompt."];
[Column 6, lines 1-5 "In another aspect, the prompt generator 128 may associate one or more pre-configured prompt templates with at least a portion of the intent and task request and populate each prompt template to generate one or more prompts accordingly."];
[Column 7, lines 18-21 "Once the additional context/information is gathered, an updated prompt will be generated by the prompt generator 128 and the ML model may then process the updated prompt based on its own request."];
[Column 6, lines 19-21 "The trained ML model may be utilized to process the intent and task objective and output one or more prompts responsive to the input" where the ML model trained to output prompts corresponds to the AI software application, and processing the intent and task objective to output one or more prompts responsive to the input corresponds to generating the first adjusted prompt utilizing the AI software application].
It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Li in view of Gray in view of Scherle in view of Beauchamp with the teachings of Schillace because utilizing AI software to generate an adjusted prompt based on the provided input and task objective, would automate the generation of a prompt adapted to the particular IDR type. The combination would advantageously reduce manual prompt engineering while producing more relevant prompts for processing different type of IDR’s.
Regarding claim 3, Li further teaches
the device of claim 1, wherein the operations comprise, prior to
receiving the IDR from the computing device associated with the first user:
identifying a group of IDR types;
[Column 10, 11 lines 63- 67, line 1 "The system may be configured to apply one or more data processing algorithms, models, and/or machine learning models (including, e.g., a sequence of machine learning techniques) to identify document types for document bundles, for single documents, and/or for single pages of documents"].
However, Li in view of Gray fails to expressly recite providing the group of IDR types to a third computing device associated with a third user;
But Scherle recites providing the group of IDR types to a third computing device associated with a third user;
[0113 "Once the data classification system 128 has obtained classification results 506 for the specific set of attributes 504, the classification results 506 are presented in the user interface 502. For example, the attributes 504 may be presented together with their corresponding classifications in a “side-by-side” manner. In the diagram 500 of FIG. 5 , the user interface 502 also presents an AI classification review section 508 that enables the user 108 to review and validate or adjust the classification results 506. Specifically, in FIG. 5 , the AI classification review section 508 provides, for each attribute, a user-selectable confirmation element 510 and a user-selectable classification adjustment element 512" where classification results indicates a group of IDR types].
It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Li in view of Gray with Scherle’s technique of providing a group of IDR’s to a computing device associated with another user and receiving a user input in response, because doing so would enable the multiple IDR’s and its associated type information to be communicated to an appropriate user for review and further action. The combination would advantageously facilitate collaborative processing of IDR’s by allowing relevant IDR information to be distributed to the appropriate user.
Li in view of Gray and in further view Scherle do not teach receiving second user-generated input from the third computing device associated with the third user, wherein the second user-generated input indicates a group of prompts; providing a series of prompts to the AI software application, wherein each of the series of prompts includes each prompt from the group of prompts and a second indication, wherein the second indication indicates to adjust each prompt based on each IDR type from the group of IDR types; and generating the group of adjusted prompts utilizing the AI software application.
However, Beauchamp teaches receiving second user-generated input from the third computing device associated with the third user, wherein the second user-generated input indicates a group of prompts;
[0083 "If multiple users (using respective multiple user devices) are involved in composing and/or editing the text, each user may independently access the text-editor 550 (e.g., using multiple instances of the text-editor 550) and/or may independently access services provided by the text-editor 550, to independently compose and/or edit the text." where multiple users indicates a third user and based on user inputs, multiple prompts can be generated];
[0086 "In this example, each user element 622 is provided with a single-instruction submit option 624 and a single-instruction cancel option 626. Selection of the single-instruction submit option 624 may cause the instruction contained in the user element 622 to be provided to the prompt generator 500 together with the corresponding selected portion of text 614 … This may enable generation of a prompts to a LLM"];
[0109 "In some examples, the prompt generator 500 may generate a series of prompts to the LLM, using a technique referred to as prompt chaining. Conceptually, prompt chaining may involve providing a series of prompts that enable the LLM to generate a desired output step-by-step"].
It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Li in view of Gray in view of Scherle with the teachings of Beauchamp because doing so would allow the user receiving the group of IDR’s to provide a group of prompts for processing the different IDR types. This combination would advantageously enable multiple IDR categories to be handled using respective user-directed prompts, thereby improving processing efficiency and flexibility.
Li in view of Gray and in view Scherle and in further view of Beauchamp do not teach providing a series of prompts to the AI software application, wherein each of the series of prompts includes each prompt from the group of prompts and a second indication, wherein the second indication indicates to adjust each prompt based on each IDR type from the group of IDR types; and generating the group of adjusted prompts utilizing the AI software application.
But Schillace teaches providing a series of prompts to the AI software application, wherein each of the series of prompts includes each prompt from the group of prompts and a second indication, wherein the second indication indicates to adjust each prompt based on each IDR type from the group of IDR types;
[Column 5, lines 13- 20 "The prompt generator 128, receives the task objective and task request and utilizes them to generate one or more prompts for the ML model. The prompt generator 128 may generate one or more prompts that when processed a ML model, such as a generative large language model (LLM), provide sufficient context for the ML model to generate model output responsive to the task objective associated with the input" where the task request can correspond to the first prompt, the task objective reflecting IDR type(context associated with the task objective) corresponds to the first indication because the task objective specifies the context to be incorporated into the prompt, the context represented by the task objective is based on the first IDR type identified by Li's reference];
[Column 5, lines 24- 30 "Thus, the one or more prompts encompass the semantic context of the task objective and task request so that the ML model can generate model output responsive to the requested task and/or intent without requiring additional training or fine-tuning of the model prior to generating model output responsive to the task or intent"];
[Column 9, lines 25- 31 "At operation 204, a task objective may be determined, for example, by a task objective module (e.g., task objective module 126) based on the input and task request. The task objective may encapsulate the general intent or specific meaning of the input and may be utilized to assist in generating one or more prompts specifically related to that objective" where task objective is determined based on the input and task request and input could be the document request from Li's reference];
[Column 5, lines 55-67 "The prompts may be generated in a variety of ways. In one example, an application and/or ML model from model repository 130 may analyze the task objective and input to select from one or more prompt templates stored in data store 106 with which to populate the prompt. In an alternative example, NL processing tools may be utilized to analyze the input and task objective to determine one or more associated prompt templates and/or populate a prompt. In a further example, the prompt generator 128 may use the ML model directly to either write a prompt, rewrite a prompt, and/or expand a seed prompt to determine one or more associated prompt templates and/or populate a prompt"].
and generating the group of adjusted prompts utilizing the AI software application
[Column 5 , lines 63-67 "In a further example, the prompt generator 128 may use the ML model directly to either write a prompt, rewrite a prompt, and/or expand a seed prompt to determine one or more associated prompt templates and/or populate a prompt."];
[Column 6, lines 1-5 "In another aspect, the prompt generator 128 may associate one or more pre-configured prompt templates with at least a portion of the intent and task request and populate each prompt template to generate one or more prompts accordingly"];
[Column 6, lines 19-21 "The trained ML model may be utilized to process the intent and task objective and output one or more prompts responsive to the input" where the ML model trained to output prompts corresponds to the AI software application, and processing the intent and task objective to output one or more prompts responsive to the input corresponds to generating the first adjusted prompt utilizing the AI software application].
It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Li in view of Gray in view of Scherle in view of Beauchamp with the teachings of Schillace because doing so would enable each prompt in a group of prompts to be adapted according to a corresponding IDR type. The combination would advantageously produce type-specific prompts for multiple IDR categories, thereby improving the relevance and effectiveness of AI processing across different types of IDR’s.
Regarding claim 6, the rejection of claim 1 is incorporated. Li in view of Gray teach all the limitations of the current invention as stated above.
Schillace further teaches the device of claim 1, wherein the generating of the response comprises generating the response with a template associated with the first IDR type.
[Column 6, lines 1-5 "In another aspect, the prompt generator 128 may associate one or more pre-configured prompt templates with at least a portion of the intent and task request and populate each prompt template to generate one or more prompts accordingly."].
It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Li in view of Gray in view of Scherle in view of Beauchamp with the teachings of Schillace because doing so would enable the AI-generated response to conform to a predefined structure appropriate for the IDR being processed. The combination would advantageously provide more consistent, organized and standardized responses, thereby improving the usability and of AI generated responses for different IDR’s.
Regarding claim 8, the rejection of claim 1 is incorporated. Li in view of Gray teach all the limitations of the current invention as stated above.
Schillace further teaches the device of claim 1, wherein the operations comprise storing
a group of previously generated IDRs in a vector database.
[Column 5, lines 46- 51 "For example, data store 106 may include one or more embeddings associated with previously generated model output and/or previously processed input, thereby enabling semantic retrieval of the prompt template and associated context (e.g., such that previously generated model output may be iterated upon)."].
It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Li in view of Gray in view of Scherle in view of Beauchamp with the teachings of Schillace because doing so would enable prior IDR’s to be efficiently retrieved based on similarity and used in processing subsequent IDR’s. This combination would advantageously leverage information from previously processed IDR’s to provide relevant contextual information for AI processing, thereby improving the accuracy and relevance of responses generated for newly received IDR’s.
Regarding claim 9, the rejection of claim 1 is incorporated. Li in view of Gray teach all the limitations of the current invention as stated above.
Schillace further teaches the device of claim 8, wherein the generating of the response
associated with the IDR comprises: accessing a portion of the group of previously generated IDRs from the vector database; and
generating the response associated with the IDR based on the portion of the group of previously generated IDRs.
[Column 5, lines 46- 51 "For example, data store 106 may include one or more embeddings associated with previously generated model output and/or previously processed input, thereby enabling semantic retrieval of the prompt template and associated context (e.g., such that previously generated model output may be iterated upon)."];
[Column 3, lines 42-49 “Alternatively, input may be programmatically generated by application 104, may be based on the content of a file or an electronic communication, may comprise an image, other data type, and/or a plurality of other examples which will be understood by one
having skill in the art. In some instances, the input may reference previously created entities or known entities ( e.g., as may be stored within data store 106).”].
It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Li in view of Gray in view of Scherle in view of Beauchamp with the teachings of Schillace because doing so would allow previously processed IDR’s having relevant information to provide contextual support for processing the current IDR. The combination would advantageously use relevant historical IDR information to generate more accurate, contextually informed, and consistent responses while avoiding the need to process unrelated previously stored IDR’s.
Claim [ 7 ] are rejected under 35 U.S.C. 103 as being unpatentable over Li (US12217318, hereinafter Li) in view of Gray (US11769017B1, hereinafter Gray) and in further view of Deutsch (US12051205B1, hereinafter Deutsch).
Regarding claim 7, the rejection of claim 1 is incorporated. Li and Gray teach all the limitations of the current invention as stated above.
However, Li and Gray do not teach the device of claim 1, wherein the operations comprise determining that the IDR includes a group of images resulting in a second determination, wherein the selecting of the first AI model comprises selecting the first AI models from the group of AI models based on the second determination resulting in the first AI model selection.
But Deutsch teaches
determining that the IDR includes a group of images resulting in a second determination, wherein the selecting of the first AI model comprises selecting the first AI models from the group of AI models based on the second determination resulting in the first AI model selection.
[Column 23, lines 17-25 "ML modeling engine 1130 may include model selector engine 1132 (e.g., configured to select a model from among a plurality of models, such as based on input data), parameter selector engine 1134 (e.g., configured to add, remove, and/or change one or more parameters of a model), and/or model generation engine 1136 (e.g., configured to generate one or more machine learning models, such as according to model input data, model output data, comparison data, and/or validation data)."];
[Column 22, lines 37-41 “System 1100 may include data input engine 1110 that can further include data retrieval engine 1104 and data transform engine 1106. Data input engine 1110 may be configured to access, interpret, request, format, re-format, or receive input
data from data source(s) 1102”];
[Column 8, lines 50-53 "In some embodiments, input data generated by the at least one processor can comprise the image, the textual prompt of the user, and a prepended textual prompt generated by, for example, the model through prompt engineering"];
[Column 6, lines 11-14 "To respond to such requests, the machine learning model may employ image processing system 120 when processing images according to the requests submitted by the user" where processing images implies multiple images].
It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Li in view of Gray with the teachings of Deutsch because doing so would enable an appropriate AI model to be selected according to the content of Li’s IDR, including when the IDR comprises image content. This combination would advantageously permit image-containing IDR’s to be processed by a model suited to the input data, thereby improving the accuracy and effectiveness of AI processing for IDR’s containing different types of content.
Claim [10, 18 ] are rejected under 35 U.S.C. 103 as being unpatentable over Li (US12217318, hereinafter Li) and in view of Scherle (US20250173550A1, hereinafter Scherle) and in further view of Beauchamp ( US20240256762A1, hereinafter Beauchamp) and in further view of Schillace (US12505296B2, hereinafter Schillace).
Regarding claim 10, Li teaches
A non-transitory machine-readable medium, comprising executable instructions that, when executed by a processing system including a processor, facilitate performance of operations, the operations comprising:
[Column 2, lines 57-60 "In some embodiments, a first non-transitory computer-readable storage medium is provided, the first non-transitory computer-readable storage medium storing instructions for determining the composition of document bundles, the instructions configured to be executed by one or more processors of a system to cause the system to: receive first input data comprising a document bundle;"];
identifying a first IDR type;
[Column 9, lines 60- 63 "In some embodiments, a system for automated document processing may be configured to perform automated document classification (e.g., classifying documents according to different document types) and/or document bundling" where determining the document type(IDR type) associated with the received document is the first determination].
However, Li fails to expressly recite providing the first IDR type to a first computing device associated with a first user;
But Scherle recites providing the first IDR type to a first computing device associated with a first user;
[0089 "At operation 318 of the method 300, the classification results are presented in a user interface using the user interface component 212 of the data classification system 128. For example, the user 108 may be presented with the classification results for review or validation in a user interface provided at the user device 106" where classification results indicates the type of request and user indicates a first user];
[0166 “The applications 720 may utilize built in operating system functions (e.g., kernel 728, services 730 or drivers 732), libraries (e.g., system libraries 734, API libraries 736, and other libraries 738), and frameworks/middleware layer 718 to create user interfaces to interact with users of the system” where user interfaces to interact with users indicates multiple users];
[0022 "The classification scheme may be used to guide or indicate to a machine learning model how to classify the process metadata. For example, the classification scheme may indicate that each attribute should be classified as either being a personal data attribute or not being a personal data attribute. Alternatively, or additionally, the classification scheme may indicate that each attribute should be classified into one of a set of predetermined categories (e.g., “text,” “numerical,” “categorical,” or “date”). The classification scheme may provide a standardized taxonomy to categorize items using consistent labels across different automated processes (e.g., within an organization). The instruction may thus include or reference the predetermined classification scheme"].
It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Li iwith Scherle’s technique of providing an IDR to a computing device associated with another user and receiving a user input in response, because doing so would enable the identified IDR and its associated type information to be communicated to an appropriate user for review and further action. The combination would advantageously facilitate collaborative processing of IDR’s by allowing relevant IDR information to be distributed to the appropriate user.
Li in view Scherle do not teach receiving first user-generated input from the first computing device associated with the first user, wherein the first user-generated input indicates a first prompt; providing a prompt to an artificial intelligence (AI) software application, wherein the prompt includes the first prompt and a first indication, wherein the first indication indicates to adjust the first prompt based on the first IDR type; and generating a first adjusted prompt utilizing the AI software application.
However, Beauchamp teaches receiving first user-generated input from the first computing device associated with the first user, wherein the first user-generated input indicates a first prompt;
[0086 "In this example, each user element 622 is provided with a single-instruction submit option 624 and a single-instruction cancel option 626. Selection of the single-instruction submit option 624 may cause the instruction contained in the user element 622 to be provided to the prompt generator 500 together with the corresponding selected portion of text 614 … This may enable generation of a first prompt based on user actions"];
[0083 "If multiple users (using respective multiple user devices) are involved in composing and/or editing the text, each user may independently access the text-editor 550 (e.g., using multiple instances of the text-editor 550) and/or may independently access services provided by the text-editor 550, to independently compose and/or edit the text." where respective multiple devices indicates multiple users which corresponds to a second user];
[0015 "In an example of any of the preceding example systems, the processing unit may be configured to execute computer-readable instructions to further cause the system to: generate the prompt including at least the portion of the annotated block of text, the prompt also including an instruction to cause the LLM to further annotate the annotated block of text in accordance with at least one inserted text-editing instruction…."]
It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Li in view of Scherle with the teachings of Beauchamp because doing so would allow the user receiving the IDR information to provide a prompt that is subsequently adapted in view of the information associated with the IDR type. This combination would advantageously provide prompts tailored to the particular type of IDR being processed, therefore would improve the relevancy and effectiveness of subsequent AI processing.
Li in view Scherle and in further view of Beauchamp do not teach providing a prompt to an artificial intelligence (AI) software application, wherein the prompt includes the first prompt and a first indication, wherein the first indication indicates to adjust the first prompt based on the first IDR type; and generating a first adjusted prompt utilizing the AI software application.
However, Schillace teaches providing a prompt to an artificial intelligence (AI) software application, wherein the prompt includes the first prompt and a first indication, wherein the first indication indicates to adjust the first prompt based on the first IDR type;
[Column 5, lines 12- 20 "The prompt generator 128, receives the task objective and task request and utilizes them to generate one or more prompts for the ML model. The prompt generator 128 may generate one or more prompts that when processed a ML model, such as a generative large language model (LLM), provide sufficient context for the ML model to generate model output responsive to the task objective associated with the input" where the task request can correspond to the first prompt, the task objective reflecting IDR type(context associated with the task objective) corresponds to the first indication because the task objective specifies the context to be incorporated into the prompt, the context represented by the task objective is based on the first IDR type identified by Li];
[Column 5, liens 24- 30 "Thus, the one or more prompts encompass the semantic context of the task objective and task request so that the ML model can generate model output responsive to the requested task and/or intent without requiring additional training or fine-tuning of the model prior to generating model output responsive to the task or intent"];
[Column 9, lines 25- 31 "At operation 204, a task objective may be determined, for example, by a task objective module (e.g., task objective module 126) based on the input and task request. The task objective may encapsulate the general intent or specific meaning of the input and may be utilized to assist in generating one or more prompts specifically related to that objective" where task objective is determined based on the input and task request and input could be the document request from Li's reference];
[Column 5, lines 55-67 "The prompts may be generated in a variety of ways. In one example, an application and/or ML model from model repository 130 may analyze the task objective and input to select from one or more prompt templates stored in data store 106 with which to populate the prompt. In an alternative example, NL processing tools may be utilized to analyze the input and task objective to determine one or more associated prompt templates and/or populate a prompt. In a further example, the prompt generator 128 may use the ML model directly to either write a prompt, rewrite a prompt, and/or expand a seed prompt to determine one or more associated prompt templates and/or populate a prompt"]
generating the first adjusted prompt utilizing the AI software application.
[Column 9, lines 63-67 "In a further example, the prompt generator 128 may use the ML model directly to either write a prompt, rewrite a prompt, and/or expand a seed prompt to determine one or more associated prompt templates and/or populate a prompt."];
[Column 6, lines 1-5 "In another aspect, the prompt generator 128 may associate one or more pre-configured prompt templates with at least a portion of the intent and task request and populate each prompt template to generate one or more prompts accordingly."];
[Column 7, lines 18-21 "Once the additional context/information is gathered, an updated prompt will be generated by the prompt generator 128 and the ML model may then process the updated prompt based on its own request."];
[Column 6, lines 19-21 "The trained ML model may be utilized to process the intent and task objective and output one or more prompts responsive to the input" where the ML model trained to output prompts corresponds to the AI software application, and processing the intent and task objective to output one or more prompts responsive to the input corresponds to generating the first adjusted prompt utilizing the AI software application].
It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Li in view of Scherle in view of Beauchamp with the teachings of Schillace because utilizing AI software to generate an adjusted prompt based on the provided input and task objective, would automate the generation of a prompt adapted to the particular IDR type. The combination would advantageously reduce manual prompt engineering while producing more relevant prompts for processing different type of IDR’s.
Regarding claim 18, Li teaches
A method, comprising: identifying, by a processing system including a processor, a group of IDR types;
[Column 9, lines 28-30 “Disclosed herein are systems and methods for providing
AI-augmented auditing platforms, including techniques for automated document processing.”];
[Column 10, 11 lines 63- 67, line 1 "The system may be configured to apply one or more data processing algorithms, models, and/or machine learning models (including, e.g., a sequence of machine learning techniques) to identify document types for document bundles, for single documents, and/or for single pages of documents"].
However, Li fails to expressly recite providing, by the processing system, the group of IDR types to a first computing device associated with a first user;
But Scherle recites providing, by the processing system, the group of IDR types to a first computing device associated with a first user;
[0113 "Once the data classification system 128 has obtained classification results 506 for the specific set of attributes 504, the classification results 506 are presented in the user interface 502. For example, the attributes 504 may be presented together with their corresponding classifications in a “side-by-side” manner. In the diagram 500 of FIG. 5 , the user interface 502 also presents an AI classification review section 508 that enables the user 108 to review and validate or adjust the classification results 506. Specifically, in FIG. 5 , the AI classification review section 508 provides, for each attribute, a user-selectable confirmation element 510 and a user-selectable classification adjustment element 512" where classification results indicates a group of IDR types].
It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Li with Scherle’s technique of providing a group of IDR’s to a computing device associated with another user and receiving a user input in response, because doing so would enable the multiple IDR’s and its associated type information to be communicated to an appropriate user for review and further action. The combination would advantageously facilitate collaborative processing of IDR’s by allowing relevant IDR information to be distributed to the appropriate user.
Li in view of Scherle do not teach receiving, by the processing system, first user-generated input from the first computing device associated with the first user, wherein the first user-generated input indicates a group of prompts; providing, by the processing system, a series of prompts to an artificial intelligence (AI) software application, wherein each of the series of prompts includes each prompt from the group of prompts and a first indication, wherein the first indication indicates to adjust each prompt based on each IDR type from the group of IDR types; and generating, by the processing system, a group of adjusted prompts utilizing the AI software application.
However, Beauchamp teaches receiving, by the processing system, first user-generated input from the first computing device associated with the first user, wherein the first user-generated input indicates a group of prompts;
[0083 "If multiple users (using respective multiple user devices) are involved in composing and/or editing the text, each user may independently access the text-editor 550 (e.g., using multiple instances of the text-editor 550) and/or may independently access services provided by the text-editor 550, to independently compose and/or edit the text." where multiple users indicates a third user and based on user inputs, multiple prompts can be generated];
[0086 "In this example, each user element 622 is provided with a single-instruction submit option 624 and a single-instruction cancel option 626. Selection of the single-instruction submit option 624 may cause the instruction contained in the user element 622 to be provided to the prompt generator 500 together with the corresponding selected portion of text 614 … This may enable generation of a prompts to a LLM"];
[0109 "In some examples, the prompt generator 500 may generate a series of prompts to the LLM, using a technique referred to as prompt chaining. Conceptually, prompt chaining may involve providing a series of prompts that enable the LLM to generate a desired output step-by-step"].
It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Li in view of Scherle with the teachings of Beauchamp because doing so would allow the user receiving the group of IDR’s to provide a group of prompts for processing the different IDR types. This combination would advantageously enable multiple IDR categories to be handled using respective user-directed prompts, thereby improving processing efficiency and flexibility.
Li in view Scherle and in further view of Beauchamp do not teach providing, by the processing system, a series of prompts to an artificial intelligence (AI) software application, wherein each of the series of prompts includes each prompt from the group of prompts and a first indication, wherein the first indication indicates to adjust each prompt based on each IDR type from the group of IDR types; and generating, by the processing system, a group of adjusted prompts utilizing the AI software application.
But Schillace teaches providing, by the processing system, a series of prompts to an artificial intelligence (AI) software application, wherein each of the series of prompts includes each prompt from the group of prompts and a first indication, wherein the first indication indicates to adjust each prompt based on each IDR type from the group of IDR types;
[Column 5, lines 13- 20 "The prompt generator 128, receives the task objective and task request and utilizes them to generate one or more prompts for the ML model. The prompt generator 128 may generate one or more prompts that when processed a ML model, such as a generative large language model (LLM), provide sufficient context for the ML model to generate model output responsive to the task objective associated with the input" where the task request can correspond to the first prompt, the task objective reflecting IDR type(context associated with the task objective) corresponds to the first indication because the task objective specifies the context to be incorporated into the prompt, the context represented by the task objective is based on the first IDR type identified by Li's reference];
[Column 5, liens 24- 30 "Thus, the one or more prompts encompass the semantic context of the task objective and task request so that the ML model can generate model output responsive to the requested task and/or intent without requiring additional training or fine-tuning of the model prior to generating model output responsive to the task or intent"];
[Column 9, lines 25- 31 "At operation 204, a task objective may be determined, for example, by a task objective module (e.g., task objective module 126) based on the input and task request. The task objective may encapsulate the general intent or specific meaning of the input and may be utilized to assist in generating one or more prompts specifically related to that objective" where task objective is determined based on the input and task request and input could be the document request from Li's reference];
[Column 5, lines 55-67 "The prompts may be generated in a variety of ways. In one example, an application and/or ML model from model repository 130 may analyze the task objective and input to select from one or more prompt templates stored in data store 106 with which to populate the prompt. In an alternative example, NL processing tools may be utilized to analyze the input and task objective to determine one or more associated prompt templates and/or populate a prompt. In a further example, the prompt generator 128 may use the ML model directly to either write a prompt, rewrite a prompt, and/or expand a seed prompt to determine one or more associated prompt templates and/or populate a prompt"].
and generating, by the processing system, a group of adjusted prompts utilizing the AI software application.
[Column 5 , lines 63-67 "In a further example, the prompt generator 128 may use the ML model directly to either write a prompt, rewrite a prompt, and/or expand a seed prompt to determine one or more associated prompt templates and/or populate a prompt."];
[Column 6, lines 1-5 "In another aspect, the prompt generator 128 may associate one or more pre-configured prompt templates with at least a portion of the intent and task request and populate each prompt template to generate one or more prompts accordingly"];
[Column 6, lines 19-21 "The trained ML model may be utilized to process the intent and task objective and output one or more prompts responsive to the input" where the ML model trained to output prompts corresponds to the AI software application, and processing the intent and task objective to output one or more prompts responsive to the input corresponds to generating the first adjusted prompt utilizing the AI software application].
It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Li in view of Scherle in view of Beauchamp with the teachings of Schillace because doing so would enable each prompt in a group of prompts to be adapted according to a corresponding IDR type. The combination would advantageously produce type-specific prompts for multiple IDR categories, thereby improving the relevance and effectiveness of AI processing across different types of IDR’s.
Claim [ 11, 12, 13, 15, 16, 17, 19, 20 ] are rejected under 35 U.S.C. 103 as being unpatentable over Li (US12217318, hereinafter Li) and in view of Scherle (US20250173550A1, hereinafter Scherle) and in further view of Beauchamp ( US20240256762A1, hereinafter Beauchamp) and in further view of Schillace (US12505296B2, hereinafter Schillace).and in further view of in view of Gray (US11769017B1, hereinafter Gray).
Regarding claim 11, the rejection of claim 10 is incorporated. Li , Scherle, Beauchamp and Schillace teach all the limitations of the current invention as stated above.
Li teaches
The non-transitory machine-readable medium of claim 10, wherein the operations comprise:
[Column 2, lines 57-60 "In some embodiments, a first non-transitory computer-readable storage medium is provided, the first non-transitory computer-readable storage medium storing instructions for determining the composition of document bundles, the instructions configured to be executed by one or more processors of a system to cause the system to: receive first input data comprising a document bundle;"];
receiving, over a communication network, an information document request (IDR) from a second computing device associated with a second user;
[Column 9, lines 37-40 "As described herein, a system for providing AI-augmented auditing platforms may be configured to receive one or more documents as input data and to perform automated processing of the input documents" where one or more documents correspond to Information document request];
[Column 10, lines 55-59 “Context data may be received by the system from any one or more suitable data sources, may be indicated by one or more user inputs detected by the system, and/or may be deduced by one or more data processing models of the system”].
determining that the IDR is associated with the first IDR type resulting in a first determination;
[Column 9, lines 60- 63 "In some embodiments, a system for automated document processing may be configured to perform automated document classification (e.g., classifying documents according to different document types) and/or document bundling" where determining the document type(IDR type) associated with the received document is the first determination].
Providing the IDR
[Column 9, lines 37-40 "As described herein, a system for providing AI-augmented auditing platforms may be configured to receive one or more documents as input data and to perform automated processing of the input documents" where one or more documents correspond to Information document request].
However, Li in view of Scherle in view of Beauchamp and in futher view of Schillace fails to expressly selecting the first adjusted prompt from a group of adjusted prompts based on the first determination; selecting a first AI model from a group of AI models based on the first determination resulting in a first AI model selection; providing the IDR, the first adjusted prompt, and the first AI model selection to the AI software application; generating a response associated with the IDR utilizing the AI software application based on the IDR, the first adjusted prompt, and the first AI model selection; and providing the response associated with the IDR to the second computing device associated with the second user.
But Gray teaches selecting the first adjusted prompt from a group of adjusted prompts based on the first determination;
[Column 5, lines 1-17 "For example, if it is determined that the user is familiar with “X content”, a prompt of “assuming the user is familiar with [description of the certain content] answer [query]” can be processed using the LLM in generating the NL based summary. In contrast, if it is determined that the user is not familiar with any content responsive to the query, a prompt of “answer [query]” can be processed using the LLM in generating the NL based summary. As another example, if it is determined that the user is familiar with “X content”, a prompt of “assuming the user is familiar with [description of the certain content] summarize [search result document content]” can be processed using the LLM in generating the NL based summary. In contrast, if it is determined that the user is not familiar with any content responsive to the query, a prompt of “summarize [search result document content]” can be processed using the LLM in generating the NL based summary” where determining condition corresponds to determination and selecting one prompt vs another implies selecting a first adjusted prompt from multiple prompts];
[Column 23, lines 38-46 "At sub-block 460A, the system generates the revised NL based summary using the same LLM as used in block 454, but using a revised input with a revised prompt that reflects familiarity with the content of the SRD(s) interacted with. For example, the prompt used in block 454 could be “create a summary of the following”, whereas the prompt used in sub-block 460A could be “create a summary of the following and assuming the user already knows X”, where “X” is a description of the content of the SRD(s) interacted with. In some implementations, the system can generate the description “X” to conform to some or all of the content of the SRD(s) interacted with"]
selecting a first artificial intelligence (AI) model from a group of AI models based on the first determination resulting in a first AI model selection;
[Column 9, 10, lines 64-67, lines 1-3 "The LLM selection engine 132 can, in response to receiving a query, determine which, if any, of multiple generative model(s) (LLM(s) and/or other generative model(s)) to utilize in generating response(s) to render responsive to the query" where multiple generative models correspond to the group of AI models, the determination of which generative model to utilize corresponds to selecting the first AI model, and the determined generative model corresponds to the first AI model selection.]
[Column 29, lines 51- 55 "Yet another example, a first prompt can be processed (along with additional content) using a first LLM and a second prompt can be processed (optionally along with the same additional content) using a second LLM" where the additional content is a third component, the first prompt corresponds to the first adjusted prompt, and the first LLM corresponds to the first AI model selection];
[Column 37, lines 44-46 "In some implementations, the candidate generative models include a given large language model (LLM) paired with a first prompt to be used in response generation…"];
[Column 9, lines 20-25 "Further, the client device 110 and/or the NL based response system 120 can include one or more memories for storage of data and/or software applications, one or more processors for accessing data and executing the software applications, and/or other components that facilitate communication over one or more of the networks 199."];
[Column 9, lines 64-66, Column 10 lines 1 "The LLM selection engine 132 can, in response to receiving a query, determine which, if any, of multiple generative model(s) (LLM(s) and/or other generative model(s)) to utilize in generating response(s) to render responsive to the query"];
generating a response associated with the IDR utilizing the AI software application
[Column 10, lines 8- "The LLM input engine 134 can, in response to receiving a query, generate LLM input that is to be processed using an LLM in generating an NL based response to the query. As described herein, such content can include query content that is based on the query and/or additional content" where additional content using the LLM to generate the NL based summary corresponds to the third component];
[Column 37, lines 44-46 "In some implementations, the candidate generative models include a given large language model (LLM) paired with a first prompt to be used in response generation…"];
[Column 23, lines 58- 62 "At sub-block 4606, the system generates the revised NL based summary using an additional LLM, relative to the one used in block 454, that is fine-tuned based on a prompt that reflects familiarity with content of the SRD(s) interacted with"];
[Column 10, lines 17-19 "The LLM response generation engine 136 can process LLM input, that is generated by the LLM input engine 134, using an LLM to generate an NL based summary"];
[Column 9, lines 20-25 "Further, the client device 110 and/or the NL based response system 120 can include one or more memories for storage of data and/or software applications, one or more processors for accessing data and executing the software applications, and/or other components that facilitate communication over one or more of the networks 199."].
providing, over the communication network, the response associated with the IDR to the second computing device associated with the second user.
[Column 32, lines 31-34 "The method further includes generating a natural language (NL) based summary using the LLM output, and causing the NL based summary to be rendered at the client device in response to the query" where the NL based summary is the response and client device is the second computing device].
It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Li with Gray because Gray teaches that different generative models may perform better in different situations and enables use of the output from the model best suited for the current situation. This combination would allow Li’s classified audit documents to be processed using an appropriately selected AI model and prompt, thereby improving the accuracy and suitability of AI generated response for different IDR types and content.
Regarding claim 12, the rejection of claim 11 is incorporated. Claim 12 is substantially the as claim 4 and is therefore rejected under the same rationale as above.
Regarding claim 13, the rejection of claim 12 is incorporated. Claim 13 is substantially the as claim 5 and is therefore rejected under the same rationale as above.
Regarding claim 14, the rejection of claim 11 is incorporated. Claim 14 is substantially the as claim 6 and is therefore rejected under the same rationale as above.
Regarding claim 16, the rejection of claim 11 is incorporated. Claim 16 is substantially the as claim 8 and is therefore rejected under the same rationale as above.
Regarding claim 17, the rejection of claim 16 is incorporated. Claim 17 is substantially the as claim 9 and is therefore rejected under the same rationale as above.
Regarding claim 19, Li teaches
receiving, by the processing system, over a communication network, an information
document request (IDR) from a second computing device associated with a second user;
[Column 9, lines 37-40 "As described herein, a system for providing AI-augmented auditing platforms may be configured to receive one or more documents as input data and to perform automated processing of the input documents" where one or more documents correspond to Information document request];
[Column 10, lines 55-59 “Context data may be received by the system from any one or more suitable data sources, may be indicated by one or more user inputs detected by the system, and/or may be deduced by one or more data processing models of the system”].
determining, by the processing system, a first IDR type associated with the IDR resulting
in a first determination, wherein the group of IDR types comprises the first IDR type;
[Column 9, lines 60- 63 "In some embodiments, a system for automated document processing may be configured to perform automated document classification (e.g., classifying documents according to different document types) and/or document bundling" where determining the document type(IDR type) associated with the received document is the first determination from ];
[Column 10, 11 lines 63- 67, line 1 "The system may be configured to apply one or more data processing algorithms, models, and/or machine learning models (including, e.g., a sequence of machine learning techniques) to identify document types for document bundles, for single documents, and/or for single pages of documents"].
Providing an IDR
[Column 9, lines 37-40 "As described herein, a system for providing AI-augmented auditing platforms may be configured to receive one or more documents as input data and to perform automated processing of the input documents" where one or more documents correspond to Information document request].
However, Li in view of Scherle in view of Beauchamp and in further view of Schillace fail to expressly recite selecting, by the processing system, a first adjusted prompt from the group of adjusted prompts based on the first determination; selecting, by the processing system, a first AI model from a group of AI models based on the first determination resulting in a first AI model selection; providing, by the processing system, IDR, the first adjusted prompt, and the first AI model selection to the AI software application; generating, by the processing system, a response associated with the IDR utilizing the AI software application based on the IDR, the first adjusted prompt, and the first AI model selection; providing, by the processing system, over the communication network, the response associated with the IDR to the second computing device associated with the second user.
But Gray teaches selecting, by the processing system, a first adjusted prompt from the group of adjusted prompts based on the first determination;
[Column 5, lines 1-17 "For example, if it is determined that the user is familiar with “X content”, a prompt of “assuming the user is familiar with [description of the certain content] answer [query]” can be processed using the LLM in generating the NL based summary. In contrast, if it is determined that the user is not familiar with any content responsive to the query, a prompt of “answer [query]” can be processed using the LLM in generating the NL based summary. As another example, if it is determined that the user is familiar with “X content”, a prompt of “assuming the user is familiar with [description of the certain content] summarize [search result document content]” can be processed using the LLM in generating the NL based summary. In contrast, if it is determined that the user is not familiar with any content responsive to the query, a prompt of “summarize [search result document content]” can be processed using the LLM in generating the NL based summary” where determining condition corresponds to determination and selecting one prompt vs another implies selecting a first adjusted prompt.];
[Column 23, lines 38-46 "At sub-block 460A, the system generates the revised NL based summary using the same LLM as used in block 454, but using a revised input with a revised prompt that reflects familiarity with the content of the SRD(s) interacted with. For example, the prompt used in block 454 could be “create a summary of the following”, whereas the prompt used in sub-block 460A could be “create a summary of the following and assuming the user already knows X”, where “X” is a description of the content of the SRD(s) interacted with. In some implementations, the system can generate the description “X” to conform to some or all of the content of the SRD(s) interacted with"]
selecting, by the processing system, a first AI model from a group of AI models based on the first determination resulting in a first AI model selection;
[Column 9, 10, lines 64-67, lines 1-3 "The LLM selection engine 132 can, in response to receiving a query, determine which, if any, of multiple generative model(s) (LLM(s) and/or other generative model(s)) to utilize in generating response(s) to render responsive to the query" where multiple generative models correspond to the group of AI models, the determination of which generative model to utilize corresponds to selecting the first AI model, and the determined generative model corresponds to the first AI model selection.]
providing, by the processing system,
[Column 29, lines 51- 55 "Yet another example, a first prompt can be processed (along with additional content) using a first LLM and a second prompt can be processed (optionally along with the same additional content) using a second LLM" where the additional content is a third component, the first prompt corresponds to the first adjusted prompt, and the first LLM corresponds to the first AI model selection];
[Column 37, lines 44-46 "In some implementations, the candidate generative models include a given large language model (LLM) paired with a first prompt to be used in response generation…"];
[Column 9, lines 20-25 "Further, the client device 110 and/or the NL based response system 120 can include one or more memories for storage of data and/or software applications, one or more processors for accessing data and executing the software applications, and/or other components that facilitate communication over one or more of the networks 199."];
[Column 9, lines 64-66, Column 10 lines 1 "The LLM selection engine 132 can, in response to receiving a query, determine which, if any, of multiple generative model(s) (LLM(s) and/or other generative model(s)) to utilize in generating response(s) to render responsive to the query"];
generating, by the processing system, a response associated with the IDR utilizing the AI software application
[Column 10, lines 8- "The LLM input engine 134 can, in response to receiving a query, generate LLM input that is to be processed using an LLM in generating an NL based response to the query. As described herein, such content can include query content that is based on the query and/or additional content" where additional content using the LLM to generate the NL based summary corresponds to the third component];
[Column 37, lines 44-46 "In some implementations, the candidate generative models include a given large language model (LLM) paired with a first prompt to be used in response generation…"];
[Column 23, lines 58- 62 "At sub-block 4606, the system generates the revised NL based summary using an additional LLM, relative to the one used in block 454, that is fine-tuned based on a prompt that reflects familiarity with content of the SRD(s) interacted with"];
[Column 10, lines 17-19 "The LLM response generation engine 136 can process LLM input, that is generated by the LLM input engine 134, using an LLM to generate an NL based summary"];
[Column 9, lines 20-25 "Further, the client device 110 and/or the NL based response system 120 can include one or more memories for storage of data and/or software applications, one or more processors for accessing data and executing the software applications, and/or other components that facilitate communication over one or more of the networks 199."].
providing, by the processing system, over the communication network, the response associated with the IDR to the second computing device associated with the second user.
[Column 32, lines 31-34 "The method further includes generating a natural language (NL) based summary using the LLM output, and causing the NL based summary to be rendered at the client device in response to the query" where the NL based summary is the response].
It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Li with Gray because Gray teaches that different generative models may perform better in different situations and enables use of the output from the model best suited for the current situation. This combination would allow Li’s classified audit documents to be processed using an appropriately selected AI model and prompt, thereby improving the accuracy and suitability of AI generated response for different IDR types and content.
Regarding claim 20, the rejection of claim 19 is incorporated. Claim 20 is substantially the same as claim 6 and is therefore rejected under the same rationale as above.
Claim [ 15 ] are rejected under 35 U.S.C. 103 as being unpatentable over Li (US12217318, hereinafter Li) and in view of Scherle (US20250173550A1, hereinafter Scherle) and in further view of Beauchamp ( US20240256762A1, hereinafter Beauchamp) and in further view of Schillace (US12505296B2, hereinafter Schillace) and in further view of in view of Gray (US11769017B1, hereinafter Gray) and in further view of Deutsch (US12051205B1, hereinafter Deutsch).
Regarding claim 15, the rejection of claim 11 is incorporated. Claim 15 is substantially the as claim 7 and is therefore rejected under the same rationale as above.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to SHEZA ABDUL AZIZ whose telephone number is (571)272-9610. The examiner can normally be reached Monday-Friday 7:30am-5pm Alternate Fridays off.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Daniel Washburn can be reached at (571) 272-5551. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/SHEZA ABDUL AZIZ/Examiner, Art Unit 2657
/DANIEL C WASHBURN/Supervisory Patent Examiner, Art Unit 2657
Prosecution Insights