CTFR 18/122,594 CTFR 87591 Notice of Pre-AIA or AIA Status 07-03-aia AIA 15-10-aia The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA . Response to Arguments 07-38 AIA Applicant’s arguments with respect to claim s 1-29 have been considered but are moot because the arguments do not apply to any of the references being used in the current rejection. Claim Rejections - 35 USC § 103 07-06 AIA 15-10-15 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. 07-20-aia AIA The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action: A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102 of this title, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made 07-21-aia AIA Claim s 1-19, 21-27, and 29 are rejected under 35 U.S.C. 103 as being unpatentable over Raju et al. (US 2024/0029031, hereinafter Raju) in view of LaRhette et al. (US 2024/0281472, hereinafter LaRhette) . Note: Applicant’s U.S. Provisional Application No. 63/408,401 filed on 9/20/2022 does not provide adequate support for the subject matter of the independent claims 1, 10 and 22. Regarding claim 1, Raju discloses A method comprising (fig. 1-18): generating prompt comprising environment information (paragraph [0064]: a target (e.g., assigned target 354A) can comprise … functional location 382 ; paragraph [0069]: a preventive maintenance target can take the form of an object representing a preventive maintenance task … The internal representation of the task can include …, links to targets ; paragraphs [0118], [0119]: hierarchy of equipment ) , at least one example task … , and an identifier of a task to be performed (paragraph [0069]: a preventive maintenance target can take the form of an object representing a preventive maintenance task . For example, the task can be a set of instructions to be carried out on one or more targets as described herein. The internal representation of the task can include a task identifier, description of the task, task details …) by an agent present in an environment (paragraphs [0029], [0069], [0254]: a service provider or worker visiting the site to execute the job, or computer-executable instructions executed on a target processor) ; and providing the prompt to at least one machine learning process to cause the machine learning process (paragraph [0048]: one or more predicted targets for assignment to the specified header target can be predicted with a machine learning model ; paragraph [0081]: Additional features can be included in the training data (e.g., a task identifier or the like). Predictions can thus be based on the same features (e.g., a header target and a task identifier) ) to generate a new plan for performing the task, the new plan comprising a set of operations performable (paragraph [0029]: whenever a preventive maintenance order is created as part of execution of the preventive maintenance plan, the targets specified by the user are included in the preventive maintenance order ; paragraph [0074]: planning software 410 stores a maintenance plan 430 that has one or more associated maintenance tasks 450A-N ; paragraph [0119]: when a new maintenance plan is created for the system, the results of the machine learning model prediction can be filtered to remove any pieces of equipment that were installed in the past but are not part of the hierarchy anymore. For example, when generating a recommendation list, the list can be filtered to remove such targets) by the agent within the environment (paragraphs [0029], [0069], [0254]: a service provider or worker visiting the site to execute the job, or computer-executable instructions executed on a target processor). Rajo does not disclose generating a prompt comprising environment information, at least one example task expressed as computer code … providing the prompt to at least one previously trained language model to cause the at least one previously trained language model to generate a new plan . LaRhette discloses generating a prompt comprising environment information, at least one example task expressed as computer code … providing the prompt to at least one previously trained language model to cause the at least one previously trained language model to generate a new plan (fig. 1-16, paragraph [0077]: 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; paragraph [0234]: The GAI prompt 1510 is a piece of text or code that is used to instruct the GAI model 1512 towards generating a desired output (e.g., generated item 1516) … The newly generated item 1516 may be multi-modal, such as … a piece of programming code, etc.; paragraph [0238]: The generated item 1516 may be post-processed for various reasons, including improving accuracy and consistency (e.g., checking for factual errors, grammatical mistakes, or inconsistencies in style or format); enhancing quality and relevance (e.g., remove irrelevant or redundant content, improve coherence and flow, ensure that the output aligns with the intended purpose ). It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the machine learning prediction method of Raju by incorporating the pre-trained language model and code-based prompting of LaRhette. The motivation would have been to improve Raju’s task generation and prediction system by utilizing a generative AI model guided by code-based prompts, which allows for more flexible, accurate, and context-aware generation of task outputs and ensures the generated operations align with the intended purpose (LaRhette paragraph [0044], [0234], [0238]). Regarding claim 2, Raju discloses further comprising: providing, to one or more machine learning processes, assertion related information and an identifier of an assertion included in the new plan (paragraph [0072]: When used for training or prediction, an identifier can be used (e.g., a target identifier) to represent a target; paragraph [0119]: when a new maintenance plan is created for the system, the results of the machine learning model prediction can be filtered ) to cause the one or more machine learning processes to evaluate truthfulness of the assertion (paragraph [0034]: Given a header target, a machine language model can predict the most likely targets. A list of candidate targets in a recommendations list can be proposed. A confidence score or relevance factor (e.g., percentage) can be included . Thus, even targets that are unrelated in the hierarchy can be rated based on how likely they are predicted to appear. The list can be ordered by confidence score to emphasize the most likely targets . As described herein, candidates can be filtered to remove dismantled items). Regarding claim 3, Raju discloses wherein the assertion related information comprises state information related to the environment and one or more examples comprising example state information, at least one example assertion related to the example state information (paragraph [0072]: When used for training or prediction, an identifier can be used (e.g., a target identifier) to represent a target. Similarly, a class or type can be used (e.g., a target class, target type , or the like)) , and at least one corresponding result indicating truthfulness of the at least one example assertion (paragraph [0034]: Given a header target, a machine language model can predict the most likely targets. A list of candidate targets in a recommendations list can be proposed. A confidence score or relevance factor (e.g., percentage) can be included . Thus, even targets that are unrelated in the hierarchy can be rated based on how likely they are predicted to appear. The list can be ordered by confidence score to emphasize the most likely targets . As described herein, candidates can be filtered to remove dismantled items ). Regarding claim 4, Raju discloses further comprising: receiving, from the agent or a separate process, an identification of the assertion as the agent is performing the new plan (paragraph [0072]: When used for training or prediction, an identifier can be used (e.g., a target identifier) to represent a target; paragraph [0119]: when a new maintenance plan is created for the system, the results of the machine learning model prediction can be filtered ; paragraph [0254]: The innovations can be described in the context of computer-executable instructions, such as those included in program modules, being executed in a computing system on a target real or virtual processor (e.g., which is ultimately executed on one or more hardware processors) . Regarding claim 5, Raju does not disclose further comprising: after the agent has performed the new plan, evaluating effectiveness of the new plan based at least in part on a comparison of a final state of the environment and a goal state of the environment . LaRhette discloses further comprising: after the agent has performed the new plan, evaluating effectiveness of the new plan based at least in part on a comparison of a final state of the environment and a goal state of the environment (paragraph [0065]: the evaluation 214 process can evaluate the relevance and/or quality of the search results produced by the LLM (e.g., where evaluation metrics are used to assess how well the model's output matches the user's intent ); paragraph [0108]: 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 … (6) the agent's behavior is episodic (it eventually ends in a terminal state ; paragraph [0109]: (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 … (ii) S_{t+1} equals the terminal state ; paragraph [0238]: The generated item 1516 may be post-processed for various reasons, including improving accuracy and consistency (e.g., checking for factual errors, grammatical mistakes, or inconsistencies in style or format); enhancing quality and relevance (e.g., remove irrelevant or redundant content, improve coherence and flow, ensure that the output aligns with the intended purpose) ; Note: the LLM acts as an “agent” taking actions until it reaches a “terminal state” (i.e., final state), and evaluating the effectiveness of the generated output by assessing how well the output matches the user’s intent or intended purpose (i.e., goal state) ). It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the teaching of Raju by incorporating LaRhette’s method of evaluating the agent’s final state/output against the user’s intended goal state. The motivation would have been to ensure that the generated output aligns with the intended purpose, and to use that evaluation to iteratively improve the accuracy, relevance, and overall quality of the model’s generated operations (LaRhette paragraphs [0065], [0238]). Regarding claim 6, Raju discloses wherein the environment information identifies one or more actions performable by the agent and one or more objects in the environment (paragraph [0061]: FIG. 3 is a block diagram of an example internal representation 300 of a preventive maintenance plan. Such a plan can comprise one or more maintenance plan nodes 330 (or simply “plans”) that describe the maintenance and inspection tasks to be performed at maintenance objects. A maintenance task node 350 (or simply “maintenance task” or “item”) describes which maintenance task(s) should take place regularly at one or more target nodes (or simply “targets,” “technical objects,” or “objects” )) , and the at least one example task comprises at least one of the one or more actions (paragraph [0062]: maintenance task 330 could represent the task of “perform safety test .”) Regarding claim 7, Raju discloses wherein a particular example task of the at least one example task (paragraph [0138]: a maintenance plan 1110 can have one or more associated tasks 1120 ) comprises a first action set of the one or more actions, the set of operations comprises a second action set of the one or more actions, and the first action set is different from the second action set (paragraph [0063]: The maintenance operations that are defined for a maintenance task (e.g., linked to a maintenance task list ) are designated as due for the targets assigned; paragraph [0069]: a preventive maintenance target can take the form of an object representing a preventive maintenance task . For example, the task can be a set of instructions to be carried out on one or more targets). Regarding claim 8, Raju discloses wherein the first action set is performed with respect to a first object set of the one or more objects, the second action is performed with respect to a second object set of the one or more objects, and the first object set is different from the second object set (paragraph [0138]: a maintenance plan 1110 can have one or more associated tasks 1120. A task has a header target 1132 and an object list of one or more additional targets 1135A-N ) . Regarding claim 9, Rajo does not disclose wherein the at least one example task comprises at least one comment written in other than computer code. LaRhette discloses wherein the at least one example task comprises at least one comment written in other than computer code (paragraph [0099]: the model is given a prompt or instruction and simply generates a response based on its pre-trained knowledge. This generated response can be sent to the instruct-LLM 610 component, which provides a model that has been trained or tuned to follow instructions given (e.g., natural language instructions).; paragraph [0234]: The GAI prompt 1510 is a piece of text or code that is used to instruct the GAI model 1512 towards generating a desired output (e.g., generated item 1516) ; paragraph [0235]: Prompt engineering is the process of designing and crafting prompts to effectively instruct and guide a GAI model toward generating desired outputs ; Note: the prompt is used to instruct the model can be a piece of text or code, and that prompts include natural language instructions and structured text to provide context and expectations for the output ). It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the teaching of Raju by incorporating LaRhette’s use of natural language text (i.e., comments in other than computer code) within the prompt. The motivation would have been to effectively instruct and guide the generative model, ensuring that the prompt accurately conveys the task, context, and expectations for the generated output (LaRhette paragraphs [0234], [0235]). Regarding claim 10, Raju discloses A system comprising (fig. 1-18): at least one processor; and memory storing instructions that when executed by the at least one processor cause the at least one processor to (paragraph [0038]: Any of the systems herein, including the system 100, can comprise at least one hardware processor and at least one memory coupled to the at least one hardware processor) : cause at least one machine learning process trained to process language (paragraph [0037]: The machine learning model 150 is thus trained with observed (e.g., historical) header preventive maintenance task targets and preventive maintenance task targets assigned to respective of the header preventive maintenance task targets (e.g., by virtue of the header target and assigned targets both being targets of the same preventive maintenance task).) to produce a plan to comprise an assertion (paragraph [0074]: FIG. 4 is a block diagram showing an example system 400 training a machine learning model 460 for machine learning recommendation for maintenance targets and can be used in any of the examples herein. In the example, planning software 410 stores a maintenance plan 430 that has one or more associated maintenance tasks 450A-N ) , the assertion to identify at least one condition that must be satisfied for a task in the plan to be performed (paragraph [0099]: only those candidates have a confidence score over a specified threshold are included on the list ) ; and provide assertion related information to the at least one machine learning process to cause the at least one machine learning process to determine whether the at least one condition is satisfied (paragraph [0074]: FIG. 4 is a block diagram showing an example system 400 training a machine learning model 460 for machine learning recommendation for maintenance targets and can be used in any of the examples herein. In the example, planning software 410 stores a maintenance plan 430 that has one or more associated maintenance tasks 450A-N ; paragraph [0099]: only those candidates have a confidence score over a specified threshold are included on the list ) . Rajo does not disclose cause at least one machine learning process trained to process language to produce a plan based at least in part on input comprising at least one example task expressed as computer code, the plan . LaRhette discloses cause at least one machine learning process trained to process language to produce a plan based at least in part on input comprising at least one example task expressed as computer code, the plan (fig. 1-16, paragraph [0077]: 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; paragraph [0234]: The GAI prompt 1510 is a piece of text or code that is used to instruct the GAI model 1512 towards generating a desired output (e.g., generated item 1516) … The newly generated item 1516 may be multi-modal, such as … a piece of programming code, etc.; paragraph [0238]: The generated item 1516 may be post-processed for various reasons, including improving accuracy and consistency (e.g., checking for factual errors, grammatical mistakes, or inconsistencies in style or format); enhancing quality and relevance (e.g., remove irrelevant or redundant content, improve coherence and flow, ensure that the output aligns with the intended purpose ). It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the machine learning prediction method of Raju by incorporating the pre-trained language model and code-based prompting of LaRhette. The motivation would have been to improve Raju’s task generation and prediction system by utilizing a generative AI model guided by code-based prompts, which allows for more flexible, accurate, and context-aware generation of task outputs and ensures the generated operations align with the intended purpose (LaRhette paragraph [0044], [0234], [0238]). Regarding claim 11, Raju discloses further comprising: an agent (paragraph [0254]: The innovations can be described in the context of computer-executable instructions, such as those included in program modules, being executed in a computing system on a target real or virtual processor (e.g., which is ultimately executed on one or more hardware processors) to perform the plan (paragraph [0029]: whenever a preventive maintenance order is created as part of execution of the preventive maintenance plan , the targets specified by the user are included in the preventive maintenance order). Regarding claim 12, Raju discloses wherein the agent comprises an autonomous machine or a semi-autonomous machine (paragraph [0254]: The innovations can be described in the context of computer-executable instructions, such as those included in program modules, being executed in a computing system on a target real or virtual processor (e.g., which is ultimately executed on one or more hardware processors ). Regarding claim 13, Raju discloses wherein the agent is to be a virtual agent to operate within a virtual environment (paragraph [0254]: The innovations can be described in the context of computer-executable instructions, such as those included in program modules, being executed in a computing system on a target real or virtual processor (e.g., which is ultimately executed on one or more hardware processors). Regarding claim 14, Raju discloses wherein the instructions, when executed by the at least one processor, cause the at least one processor to (paragraph [0038]: Any of the systems herein, including the system 100, can comprise at least one hardware processor and at least one memory coupled to the at least one hardware processor) : receive, from the agent or a separate process, an identification of the assertion (paragraph [0072]: When used for training or prediction, an identifier can be used (e.g., a target identifier) to represent a target; paragraph [0119]: when a new maintenance plan is created for the system, the results of the machine learning model prediction can be filtered ) as the agent (paragraph [0254]: The innovations can be described in the context of computer-executable instructions, such as those included in program modules, being executed in a computing system on a target real or virtual processor (e.g., which is ultimately executed on one or more hardware processors) performs the plan (paragraph [0029]: whenever a preventive maintenance order is created as part of execution of the preventive maintenance plan , the targets specified by the user are included in the preventive maintenance order) , and the at least one processor is to provide the assertion related information (paragraph [0072]: When used for training or prediction, an identifier can be used (e.g., a target identifier) to represent a target; paragraph [0119]: when a new maintenance plan is created for the system, the results of the machine learning model prediction can be filtered ) to the at least one machine learning process (paragraph [0034]: Given a header target, a machine language model can predict the most likely targets. A list of candidate targets in a recommendations list can be proposed. A confidence score or relevance factor (e.g., percentage) can be included . Thus, even targets that are unrelated in the hierarchy can be rated based on how likely they are predicted to appear. The list can be ordered by confidence score to emphasize the most likely targets . As described herein, candidates can be filtered to remove dismantled items) after receiving the identification (paragraph [0072]: When used for training or prediction, an identifier can be used (e.g., a target identifier) to represent a target; paragraph [0119]: when a new maintenance plan is created for the system, the results of the machine learning model prediction can be filtered ). . Regarding claim 15, Raju discloses wherein the assertion related information comprises state information related to an environment in which the plan is to be performed and one or more examples comprising example state information (paragraph [0072]: When used for training or prediction, an identifier can be used (e.g., a target identifier) to represent a target. Similarly, a class or type can be used (e.g., a target class, target type , or the like)) , at least one example assertion related to the example state information, and at least one corresponding result indicating truthfulness of the at least one example assertion (paragraph [0034]: Given a header target, a machine language model can predict the most likely targets. A list of candidate targets in a recommendations list can be proposed. A confidence score or relevance factor (e.g., percentage) can be included . Thus, even targets that are unrelated in the hierarchy can be rated based on how likely they are predicted to appear. The list can be ordered by confidence score to emphasize the most likely targets . As described herein, candidates can be filtered to remove dismantled items ). Regarding claim 16, Rajo does not disclose further comprising: after the agent has performed the new plan, evaluating effectiveness of the new plan based at least in part on a comparison of a final state of the environment and a goal state of the environment . LaRhette discloses further comprising: after the agent has performed the new plan, evaluating effectiveness of the new plan based at least in part on a comparison of a final state of the environment and a goal state of the environment (paragraph [0065]: the evaluation 214 process can evaluate the relevance and/or quality of the search results produced by the LLM (e.g., where evaluation metrics are used to assess how well the model's output matches the user's intent ); paragraph [0108]: 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 … (6) the agent's behavior is episodic (it eventually ends in a terminal state ; paragraph [0109]: (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 … (ii) S_{t+1} equals the terminal state ; paragraph [0238]: The generated item 1516 may be post-processed for various reasons, including improving accuracy and consistency (e.g., checking for factual errors, grammatical mistakes, or inconsistencies in style or format); enhancing quality and relevance (e.g., remove irrelevant or redundant content, improve coherence and flow, ensure that the output aligns with the intended purpose) ; Note: the LLM acts as an “agent” taking actions until it reaches a “terminal state” (i.e., final state), and evaluating the effectiveness of the generated output by assessing how well the output matches the user’s intent or intended purpose (i.e., goal state) ). It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the teaching of Raju by incorporating LaRhette’s method of evaluating the agent’s final state/output against the user’s intended goal state. The motivation would have been to ensure that the generated output aligns with the intended purpose, and to use that evaluation to iteratively improve the accuracy, relevance, and overall quality of the model’s generated operations (LaRhette paragraphs [0065], [0238]). Regarding claim 17, Raju discloses the instructions, when executed by the at least one processor, cause the at least one processor (paragraph [0038]: Any of the systems herein, including the system 100, can comprise at least one hardware processor and at least one memory coupled to the at least one hardware processor) to perform the at least one machine learning process (paragraph [0048]: one or more predicted targets for assignment to the specified header target can be predicted with a machine learning model ; paragraph [0081]: Additional features can be included in the training data (e.g., a task identifier or the like). Predictions can thus be based on the same features (e.g., a header target and a task identifier) ). Regarding claim 18, Rajo does not disclose wherein the at least one machine learning process comprises at least one Large Language Model (“LLM”) . LaRhette discloses wherein the at least one machine learning process comprises at least one Large Language Model (“LLM”) (paragraph [0099]: the model is given a prompt or instruction and simply generates a response based on its pre-trained knowledge. This generated response can be sent to the instruct-LLM 610 component, which provides a model that has been trained or tuned to follow instructions given (e.g., natural language instructions).; paragraph [0234]: The GAI prompt 1510 is a piece of text or code that is used to instruct the GAI model 1512 towards generating a desired output (e.g., generated item 1516) ; paragraph [0235]: Prompt engineering is the process of designing and crafting prompts to effectively instruct and guide a GAI model toward generating desired outputs ; Note: the prompt is used to instruct the model can be a piece of text or code, and that prompts include natural language instructions and structured text to provide context and expectations for the output ). It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the teaching of Raju by incorporating LaRhette’s use of natural language text (i.e., comments in other than computer code) within the prompt. The motivation would have been to effectively instruct and guide the generative model, ensuring that the prompt accurately conveys the task, context, and expectations for the generated output (LaRhette paragraphs [0234], [0235]). Regarding claim 19, Rajo does not disclose wherein the LLM was trained on a corpus of text comprising text written in at least one natural language . LaRhette discloses wherein the LLM was trained on a corpus of text comprising text written in at least one natural language (paragraph [0099]: the model is given a prompt or instruction and simply generates a response based on its pre-trained knowledge. This generated response can be sent to the instruct-LLM 610 component, which provides a model that has been trained or tuned to follow instructions given (e.g., natural language instructions).; paragraph [0234]: The GAI prompt 1510 is a piece of text or code that is used to instruct the GAI model 1512 towards generating a desired output (e.g., generated item 1516) ; paragraph [0235]: Prompt engineering is the process of designing and crafting prompts to effectively instruct and guide a GAI model toward generating desired outputs ; Note: the prompt is used to instruct the model can be a piece of text or code, and that prompts include natural language instructions and structured text to provide context and expectations for the output ). It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the teaching of Raju by incorporating LaRhette’s use of natural language text (i.e., comments in other than computer code) within the prompt. The motivation would have been to effectively instruct and guide the generative model, ensuring that the prompt accurately conveys the task, context, and expectations for the generated output (LaRhette paragraphs [0234], [0235]). Regarding claim 21, Raju discloses wherein the instructions, when executed by the at least one processor, cause the at least one processor (paragraph [0038]: Any of the systems herein, including the system 100, can comprise at least one hardware processor and at least one memory coupled to the at least one hardware processor) to provide data representing one or more actions, data representing one or more objects in an environment in which the plan is to be performed (paragraph [0061]: FIG. 3 is a block diagram of an example internal representation 300 of a preventive maintenance plan. Such a plan can comprise one or more maintenance plan nodes 330 (or simply “plans”) that describe the maintenance and inspection tasks to be performed at maintenance objects. A maintenance task node 350 (or simply “maintenance task” or “item”) describes which maintenance task(s) should take place regularly at one or more target nodes (or simply “targets,” “technical objects,” or “objects” )) , and the at least one task example comprising at least one of the one or more actions to the at least one machine learning process to cause the at least one machine learning process (paragraph [0078]: In additional to observed (e.g., historical) data showing past target assigned (e.g., as currently stored in maintenance plans) , data from historical maintenance orders, historical maintenance notifications, purchase orders, bills of material, and the like can be included) to produce the plan (paragraph [0079]: the model can generate a recommendations list as described herein based on such observed, historical assignments that were made in the past ). Regarding claim 22, Raju discloses A processor comprising (fig. 1-18): one or more circuits (paragraph [0249]: processor in an application-specific integrated circuit (ASIC)) to: generate prompt to comprise operating information and an identifier of a task to be performed (paragraph [0069]: a preventive maintenance target can take the form of an object representing a preventive maintenance task . For example, the task can be a set of instructions to be carried out on one or more targets as described herein. The internal representation of the task can include a task identifier, description of the task, task details, links to targets, specified spare parts (e.g., screws, bolts, grease can, or the like), links to external services (e.g., where a service provider visits the site and executes the maintenance job on behalf of the customer) , and the like) by an agent present in an environment (paragraphs [0029], [0069], [0254]: a service provider or worker visiting the site to execute the job, or computer-executable instructions executed on a target processor) , the task to be expressed in manner that renders the task unperformable (paragraph [0059]: the score can be used to filter out those targets with low confidence scores (e.g., failing under a specified low threshold or floor )) by the agent (paragraphs [0029], [0069], [0254]: a service provider or worker visiting the site to execute the job, or computer-executable instructions executed on a target processor) ; and provide prompt to at least one machine learning process (paragraph [0048]: one or more predicted targets for assignment to the specified header target can be predicted with a machine learning model ; paragraph [0081]: Additional features can be included in the training data (e.g., a task identifier or the like). Predictions can thus be based on the same features (e.g., a header target and a task identifier) ) to cause the at least one machine learning process (paragraph [0078]: In additional to observed (e.g., historical) data showing past target assigned (e.g., as currently stored in maintenance plans) , data from historical maintenance orders, historical maintenance notifications, purchase orders, bills of material, and the like can be included) to generate a plan (paragraph [0079]: the model can generate a recommendations list as described herein based on such observed, historical assignments that were made in the past ) comprising one or more tasks (paragraph [0074]: planning software 410 stores a maintenance plan 430 that has one or more associated maintenance tasks 450A-N ) to be performable by the agent (paragraphs [0029], [0069], [0254]: a service provider or worker visiting the site to execute the job, or computer-executable instructions executed on a target processor) . Rajo does not disclose generate a prompt to comprise operating information … the operating information to comprise at least one task example expressed as computer code . LaRhette discloses generate a prompt to comprise operating information … the operating information to comprise at least one task example expressed as computer code (fig. 1-16, paragraph [0077]: 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; paragraph [0234]: The GAI prompt 1510 is a piece of text or code that is used to instruct the GAI model 1512 towards generating a desired output (e.g., generated item 1516) … The newly generated item 1516 may be multi-modal, such as … a piece of programming code, etc.; paragraph [0238]: The generated item 1516 may be post-processed for various reasons, including improving accuracy and consistency (e.g., checking for factual errors, grammatical mistakes, or inconsistencies in style or format); enhancing quality and relevance (e.g., remove irrelevant or redundant content, improve coherence and flow, ensure that the output aligns with the intended purpose ). It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the machine learning prediction method of Raju by incorporating the pre-trained language model and code-based prompting of LaRhette. The motivation would have been to improve Raju’s task generation and prediction system by utilizing a generative AI model guided by code-based prompts, which allows for more flexible, accurate, and context-aware generation of task outputs and ensures the generated operations align with the intended purpose (LaRhette paragraph [0044], [0234], [0238]). Regarding claim 23, Raju discloses wherein the one or more circuits are to (paragraph [0249]: processor in an application-specific integrated circuit (ASIC)) : receive, from the agent or a separate process, an identification of an assertion as the agent performs the plan, the assertion to identify at least one condition that must be satisfied for at least one of the one or more tasks in the plan to be performed (paragraph [0072]: When used for training or prediction, an identifier can be used (e.g., a target identifier) to represent a target; paragraph [0119]: when a new maintenance plan is created for the system, the results of the machine learning model prediction can be filtered ; paragraph [0254]: The innovations can be described in the context of computer-executable instructions, such as those included in program modules, being executed in a computing system on a target real or virtual processor (e.g., which is ultimately executed on one or more hardware processors) ; and provide assertion related information to the at least one machine learning process to cause the at least one machine learning process to determine whether the at least one condition is satisfied (paragraph [0074]: FIG. 4 is a block diagram showing an example system 400 training a machine learning model 460 for machine learning recommendation for maintenance targets and can be used in any of the examples herein. In the example, planning software 410 stores a maintenance plan 430 that has one or more associated maintenance tasks 450A-N ; paragraph [0099]: only those candidates have a confidence score over a specified threshold are included on the list ). Regarding claim 24, Raju discloses wherein the assertion related information comprises state information related to the environment and one or more examples comprising example state information (paragraph [0072]: When used for training or prediction, an identifier can be used (e.g., a target identifier) to represent a target. Similarly, a class or type can be used (e.g., a target class, target type , or the like)) , at least one example assertion related to the example state information, and at least one corresponding result indicating truthfulness of the at least one example assertion (paragraph [0034]: Given a header target, a machine language model can predict the most likely targets. A list of candidate targets in a recommendations list can be proposed. A confidence score or relevance factor (e.g., percentage) can be included . Thus, even targets that are unrelated in the hierarchy can be rated based on how likely they are predicted to appear. The list can be ordered by confidence score to emphasize the most likely targets . As described herein, candidates can be filtered to remove dismantled items ). Regarding claim 25, Raju does not disclose wherein the one or more circuits are to evaluate effectiveness of the new plan based at least in part on a comparison of a final state of the environment and a goal state of the environment . LaRhette discloses wherein the one or more circuits are to evaluate effectiveness of the new plan based at least in part on a comparison of a final state of the environment and a goal state of the environment (paragraph [0065]: the evaluation 214 process can evaluate the relevance and/or quality of the search results produced by the LLM (e.g., where evaluation metrics are used to assess how well the model's output matches the user's intent ); paragraph [0108]: 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 … (6) the agent's behavior is episodic (it eventually ends in a terminal state ; paragraph [0109]: (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 … (ii) S_{t+1} equals the terminal state ; paragraph [0238]: The generated item 1516 may be post-processed for various reasons, including improving accuracy and consistency (e.g., checking for factual errors, grammatical mistakes, or inconsistencies in style or format); enhancing quality and relevance (e.g., remove irrelevant or redundant content, improve coherence and flow, ensure that the output aligns with the intended purpose) ; Note: the LLM acts as an “agent” taking actions until it reaches a “terminal state” (i.e., final state), and evaluating the effectiveness of the generated output by assessing how well the output matches the user’s intent or intended purpose (i.e., goal state) ). It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the teaching of Raju by incorporating LaRhette’s method of evaluating the agent’s final state/output against the user’s intended goal state. The motivation would have been to ensure that the generated output aligns with the intended purpose, and to use that evaluation to iteratively improve the accuracy, relevance, and overall quality of the model’s generated operations (LaRhette paragraphs [0065], [0238]). Regarding claim 26, Raju discloses wherein the one or more circuits (paragraph [0249]: processor in an application-specific integrated circuit (ASIC)) are to perform the at least one machine learning process (paragraph [0074]: FIG. 4 is a block diagram showing an example system 400 training a machine learning model 460 for machine learning recommendation for maintenance targets and can be used in any of the examples herein. In the example, planning software 410 stores a maintenance plan 430 that has one or more associated maintenance tasks 450A-N ) . Rajo does not disclose the at least one machine learning process comprises at least one Large Language Model (“LLM”). LaRhette discloses the at least one machine learning process comprises at least one Large Language Model (“LLM”) (paragraph [0099]: the model is given a prompt or instruction and simply generates a response based on its pre-trained knowledge. This generated response can be sent to the instruct-LLM 610 component, which provides a model that has been trained or tuned to follow instructions given (e.g., natural language instructions).; paragraph [0234]: The GAI prompt 1510 is a piece of text or code that is used to instruct the GAI model 1512 towards generating a desired output (e.g., generated item 1516) ; paragraph [0235]: Prompt engineering is the process of designing and crafting prompts to effectively instruct and guide a GAI model toward generating desired outputs ; Note: the prompt is used to instruct the model can be a piece of text or code, and that prompts include natural language instructions and structured text to provide context and expectations for the output ). It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the teaching of Raju by incorporating LaRhette’s use of natural language text (i.e., comments in other than computer code) within the prompt. The motivation would have been to effectively instruct and guide the generative model, ensuring that the prompt accurately conveys the task, context, and expectations for the generated output (LaRhette paragraphs [0234], [0235]). Regarding claim 27, Rajo does not disclose wherein the LLM was trained on a corpus of text comprising text written in at least one natural language . LaRhette discloses wherein the LLM was trained on a corpus of text comprising text written in at least one natural language (paragraph [0099]: the model is given a prompt or instruction and simply generates a response based on its pre-trained knowledge. This generated response can be sent to the instruct-LLM 610 component, which provides a model that has been trained or tuned to follow instructions given (e.g., natural language instructions).; paragraph [0234]: The GAI prompt 1510 is a piece of text or code that is used to instruct the GAI model 1512 towards generating a desired output (e.g., generated item 1516) ; paragraph [0235]: Prompt engineering is the process of designing and crafting prompts to effectively instruct and guide a GAI model toward generating desired outputs ; Note: the prompt is used to instruct the model can be a piece of text or code, and that prompts include natural language instructions and structured text to provide context and expectations for the output ). It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the teaching of Raju by incorporating LaRhette’s use of natural language text (i.e., comments in other than computer code) within the prompt. The motivation would have been to effectively instruct and guide the generative model, ensuring that the prompt accurately conveys the task, context, and expectations for the generated output (LaRhette paragraphs [0234], [0235]). Regarding claim 29, Raju discloses wherein the operating information comprises data representing one or more actions performable by the agent, one or more data representing objects in the environment (paragraph [0061]: FIG. 3 is a block diagram of an example internal representation 300 of a preventive maintenance plan. Such a plan can comprise one or more maintenance plan nodes 330 (or simply “plans”) that describe the maintenance and inspection tasks to be performed at maintenance objects. A maintenance task node 350 (or simply “maintenance task” or “item”) describes which maintenance task(s) should take place regularly at one or more target nodes (or simply “targets,” “technical objects,” or “objects” )) , and … the at least one task example implements at least one of the one or more actions (paragraph [0062]: maintenance task 330 could represent the task of “perform safety test .”). Rajo does not disclose the computer code of the at least one task example . LaRhette discloses the computer code of the at least one task example (paragraph [0077]: 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; paragraph [0234]: The GAI prompt 1510 is a piece of text or code that is used to instruct the GAI model 1512 towards generating a desired output (e.g., generated item 1516) … The newly generated item 1516 may be multi-modal, such as … a piece of programming code, etc.; paragraph [0238]: The generated item 1516 may be post-processed for various reasons, including improving accuracy and consistency (e.g., checking for factual errors, grammatical mistakes, or inconsistencies in style or format); enhancing quality and relevance (e.g., remove irrelevant or redundant content, improve coherence and flow, ensure that the output aligns with the intended purpose ). It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the machine learning prediction method of Raju by incorporating the pre-trained language model and code-based prompting of LaRhette. The motivation would have been to improve Raju’s task generation and prediction system by utilizing a generative AI model guided by code-based prompts, which allows for more flexible, accurate, and context-aware generation of task outputs and ensures the generated operations align with the intended purpose (LaRhette paragraph [0044], [0234], [0238]) . 07-21-aia AIA Claim s 20 and 28 are rejected under 35 U.S.C. 103 as being unpatentable over Raju in view of LaRhette as applied to claims 19 and 27, and further in view of Miller et al. (US 2024/0289545; Provisional Application No. 63/449,035 filed on 2/28/2023 is cited for claim mapping, hereinafter Miller) . Note: Applicant’s U.S. Provisional Application No. 63/408,401 filed on 9/20/2022 does not provide adequate support for the subject matter of the dependent claims 20 and 28. Regarding claim 20, Raju in view of LaRhette does not disclose wherein both the corpus of text and the plan comprise computer code . Miller discloses wherein both the corpus of text (provisional paragraph [0036]: The AI system 120 may use various natural language processing (NLP) techniques to determine context around keywords and phrases in the text data … Th e LLM 130 in the present AI system 120 can be trained on large amounts of text data to learn the patterns and relationships between words and phrases in different contexts) and the plan (paragraph [0041]: the plan creation component 150 may leverage additional prompts 152 and responses to/from the LLM 130 itself in building a solution plan 156) comprise computer code (provisional paragraph [0053]: As illustrated the skills and resources may include any variety of web search, database, file input/output (I/0), external or internal APis, code written by the AI based engine 120 or other prompts from the AI based engine 120; paragraph [0054]: written by the AI may include any variety of source code including, but not limited to, pseudocode, code snippets, programs, and algorithms in a variety oflanguages, such as, Python, Java/JavaScript, CIC++, Ruby, PHP, SQL, Swift, React, Angular, Django, Rust, HTML/CSS, XML, to name a few). It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the teaching of Raju in view of LaRhette by incorporating Miller’s wrting, by the AI, code for large amounts of text data and solution plan. The motivation would have been to generate plans dynamically, allowing for a more comprehensive solution space with greater flexibility in executing tasks (Miller paragraph [0030]). Regarding claim 28, Raju in view of LaRhette does not disclose wherein both the corpus of text and the plan comprise computer code . Miller discloses wherein both the corpus of text (provisional paragraph [0036]: The AI system 120 may use various natural language processing (NLP) techniques to determine context around keywords and phrases in the text data … Th e LLM 130 in the present AI system 120 can be trained on large amounts of text data to learn the patterns and relationships between words and phrases in different contexts) and the plan (paragraph [0041]: the plan creation component 150 may leverage additional prompts 152 and responses to/from the LLM 130 itself in building a solution plan 156) comprise computer code (provisional paragraph [0053]: As illustrated the skills and resources may include any variety of web search, database, file input/output (I/0), external or internal APis, code written by the AI based engine 120 or other prompts from the AI based engine 120; paragraph [0054]: written by the AI may include any variety of source code including, but not limited to, pseudocode, code snippets, programs, and algorithms in a variety oflanguages, such as, Python, Java/JavaScript, CIC++, Ruby, PHP, SQL, Swift, React, Angular, Django, Rust, HTML/CSS, XML, to name a few). It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the teaching of Raju in view of LaRhette by incorporating Miller’s wrting, by the AI, code for large amounts of text data and solution plan. The motivation would have been to generate plans dynamically, allowing for a more comprehensive solution space with greater flexibility in executing tasks (Miller paragraph [0030]) . Conclusion 07-96 AIA The prior art made of record and not relied upon is considered pertinent to applicant’s disclosure. Nouri et al. (US 2024/0038226) discloses “The task prompt generator 112 generates a prompt to the user using the client computing device 102. The prompt may describe the next course of action in executing the task and/or request information associated with the task” (paragraph [0026]) and “the language model 130 represents a large natural language processing model including a deep learning model that has been trained using hundreds of millions of tokens as training data” (paragraph [0027]). Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL . See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in [0037] 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 [0037] 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 SISLEY N. KIM whose telephone number is (571)270-7832. The examiner can normally be reached M-F 11:30AM -7:30PM. 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, April Y. Blair can be reached on (571)270-1014. 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. /SISLEY N KIM/Primary Examiner, Art Unit 2196 5/30/2026 Application/Control Number: 18/122,594 Page 2 Art Unit: 2196 Application/Control Number: 18/122,594 Page 3 Art Unit: 2196 Application/Control Number: 18/122,594 Page 4 Art Unit: 2196 Application/Control Number: 18/122,594 Page 5 Art Unit: 2196 Application/Control Number: 18/122,594 Page 6 Art Unit: 2196 Application/Control Number: 18/122,594 Page 7 Art Unit: 2196 Application/Control Number: 18/122,594 Page 8 Art Unit: 2196 Application/Control Number: 18/122,594 Page 9 Art Unit: 2196 Application/Control Number: 18/122,594 Page 10 Art Unit: 2196 Application/Control Number: 18/122,594 Page 11 Art Unit: 2196 Application/Control Number: 18/122,594 Page 12 Art Unit: 2196 Application/Control Number: 18/122,594 Page 13 Art Unit: 2196 Application/Control Number: 18/122,594 Page 14 Art Unit: 2196 Application/Control Number: 18/122,594 Page 15 Art Unit: 2196 Application/Control Number: 18/122,594 Page 16 Art Unit: 2196 Application/Control Number: 18/122,594 Page 17 Art Unit: 2196 Application/Control Number: 18/122,594 Page 18 Art Unit: 2196 Application/Control Number: 18/122,594 Page 19 Art Unit: 2196 Application/Control Number: 18/122,594 Page 20 Art Unit: 2196 Application/Control Number: 18/122,594 Page 21 Art Unit: 2196 Application/Control Number: 18/122,594 Page 22 Art Unit: 2196 Application/Control Number: 18/122,594 Page 23 Art Unit: 2196 Application/Control Number: 18/122,594 Page 24 Art Unit: 2196 Application/Control Number: 18/122,594 Page 25 Art Unit: 2196 Application/Control Number: 18/122,594 Page 26 Art Unit: 2196 Application/Control Number: 18/122,594 Page 27 Art Unit: 2196
Prosecution Insights