DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Notice to Applicant
The following is a Final Office action. In response to Examiner’s Non-Final Rejection of 2/12/26, Applicant, on 5/11/26, amended claims. Claims 1-20 are pending in this application and have been rejected below.
Information Disclosure Statement
The information disclosure statement (IDS) submitted on 5/31/26 are being considered by the examiner.
Response to Amendment
Applicant’s amendments are acknowledged.
Reasons for Subject Matter Eligibility under 35 USC 101
The claim 1 overcomes the 101 rejections because the claim is now a computing system; receiving a user query; automatically identifying, by an autonomous agent, a target process that matches the user query; automatically prompting a large language model, by the autonomous agent, with a set of tools bound to the processor workflow, the user query, and a context prompt describing the target process retrieved by the autonomous agent, to determinate invocation of an API, receiving, by the autonomous agent, a response generated by the large language model; when user intervention required, dynamically adjusting a process workflow to incorporate input before invoking a specific API, when user intervention not required, automatically invoking by the autonomous agent, the specific API to execute the targe task. When viewing the claim as a whole, this when combined with the earlier limitations is viewed as “not directed to an abstract idea”; or a practical application under step 2a, prong 2, as the claim is improving another technology when viewing all the limitations listed above (See MPEP 2106.05a) and/or is viewed as a using a judicial exception in a meaningful way under MPEP 2106.05(e). Applicant’s remarks (5/11/26, page 13-14) also point to [0126-0127] where the specification explains technical aspects of autonomous agents, API, and LLMs in the claims. The same reasons also apply to claim 11 and 20 which have similar limitations.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claims 1-20 are rejected under 35 U.S.C. 103 as being unpatentable over Touati (US 2022/0130380) in view of Lu (CN 117472552).
Concerning claim 1, Touati discloses:
A computing system for improving process flow automation in an enterprise resource planning (ERP) platform (Touati – see par 54 - creation of an integration or process automation via a chatbot is illustrated in FIGS. 5A-5D; see par 80 -example of an integration flow, user 102 may build an integration scenario that synchronizes data between an enterprise resource planning (ERP) system and a customer relationship management (CRM) system; see par 90, FIG. 10 – systems 120 may provide ERP system; see par 162-163, FIG. 15 – cloud computing node is a computer system/server), the computing system comprising:
memory (Touati – see par 165 - As shown in FIG. 15, computer system/server 12 in cloud computing node 10 is shown in the form of a general-purpose computing device. The components of computer system/server 12 may include, but are not limited to, one or more processors or processing units 16, a system memory 28);
one or more hardware processors coupled to the memory (Touati – see par 165 - As shown in FIG. 15, computer system/server 12 in cloud computing node 10 is shown in the form of a general-purpose computing device. The components of computer system/server 12 may include, but are not limited to, one or more processors or processing units 16, a system memory 28); and
one or more computer readable storage media storing instructions that, when loaded into the memory, cause the one or more hardware processors to perform operations (Touati – see par 34-35 - The computer readable storage medium can be a tangible device that can retain and store instructions for use by an instruction execution device; see par 168-170 – system memory 28 includes system readable media; program/modules 42 stored in memory carry out functions/embodiments) comprising:
receiving a user query from a user interface of the ERP platform (Touati – see par 53 - The Bots expose connected systems as bots where users can perform actions through the use of speech. NLP Bots, which rely on natural language processing to interpret free text and make use of intent and entity extraction, interpretation and mapping to functional actions involving software integration. In the case of NLP bots, a user may ask natural language questions in a dialog box, i.e., to a virtual developer, and an answer may be provided back to the user based on information available in the integrated systems. If more specifics are required to formulate the answer, the user may be prompted for clarifications; see par 84, FIG. 10 - User interface components 112 may be employed by integration application design tool 110 to provide components used by integration application design tool 110 to render a user interface for view by user 102 via device 104; see par 91 - In example embodiments, a user may be able to communicate to the both in natural language and/or low- or no-code development environments as well as to ask status of the integration involving a specific component of the enterprise software or other queries of the specific integration or the global systemic ramifications thereof.);
automatically identifying, by an autonomous agent (Touati – see par 129 - FIG. 14 is a functional block diagram of an example network architecture including an example artificial intelligence (AI)-based conversational querying (CQ) platform, according to example embodiments. see par 143 - Data lake 126 may be configured to store, for example, information and data collected from CQ platform 1002 relating to automated chats (i.e., via chat bot 132); conversational details (e.g., one or more user inputs, one or more user intents identified by CQ platform 1002), a target process that matches the user query (Touati – see par 54 - creation of an integration or process automation via a chatbot is illustrated in FIGS. 5A-5D, according to various example embodiments. In FIG. 5A, the integration starts when a command 510 to, e.g., “create an integration” is entered in the chat box. In exemplary embodiments, the user may not need to enter that exact command, as the NLP algorithms can interpret the meaning of natural language. For example, a user could say “I want to create a process automation”, or “let's make an integration”, or “I want to automatically integrate ServiceNow with Jira”, and the like. see par 58 - NLP can suggest the closest matches from a library; see par 96 - With the answers to these questions, example embodiments build numerous integrations, robotic process automations, and business workflows. An example of this would be an integration to move incidents from ServiceNow to Jira. A user may say “I would like to open a bug in Jira for each urgent incident in ServiceNow in real time.” By doing so, a conversational agent may be configured to extract the source, target, event, action, and frequency of the integration, and set it up for the end user. see par 100 - Example embodiments make of intent-based ML to facilitate software integrations and hyperautomation. An intent represents the purpose of a user's input. Example embodiments include a platform to map an intent for each type of user request to the API and applications supported. Coupling entity extraction in the natural language processing (NLP) of the invention enables a term or object that is relevant to your intents and that provides a specific context for an intent.), wherein the target process is one of a plurality of processes included in a process workflow, wherein the target process includes a plurality of tasks and links connecting the plurality of tasks, wherein the links define an operation sequence of the plurality of tasks (Touati – see par 54 - In the illustrated example, the origin system is identified as “ServiceNow.” FIG. 5C illustrates the integration of files from ServiceNow to allow documents to flow from ServiceNow to another service, in this case Google Cloud Storage. see par 76 - User 102 may be a developer or other individual designing, developing, and deploying integration flows using an integration application design tool in an integration platform. User 102 may be a member of a business, organization, or other suitable group. User 102 may be a human being, but user 102 may also be an artificial intelligence construct. see par 86 - Auto-mapping recommender 116 may be leveraged by integration application design tool 110 to provide a suggested linking between fields in a source asset and fields in a target asset, i.e. an auto-mapping recommendation. see par 87 - Auto-mapping recommender 116 may suggest mappings to user 102 in a graphical representation. Such a graphical representation may demonstrate connections between the fields of one asset to the fields of the second asset with lines, arrows, or other visual connectors. Auto-mapping recommender 116 may further translate recommended mappings into a script in an expression or scripting language. Such translation is advantageous because user 102 may view an alternate, textual form of the recommendation and make modifications therein. Moreover, the script may subsequently be executed at runtime to link the source fields and target fields;
Lu also discloses “automatically identifying, by an autonomous agent, a target process that matches the user query…” – see page 8, 2nd to last paragraph - the task flow auxiliary arrangement can be performed through interaction with AI implementation, after the large model identifies the user intention, the approximate flow stored in the system is inquired through the knowledge map, the task execution API is generated to invoke the node order, according to the user intention, the parameter is automatically filled, the node is arranged for
the second time and modified in sequence through the multi-round conversation));
automatically retrieving, by the autonomous agent, a context prompt describing the target process (Touati – see par 65 - The use of full NLP can make use of a limited or predictive vocabulary that narrows the possible language ‘matches’ and increases accuracy of the NLP in the context of software integration. The bot may also recommend viable next actions to users, answer questions, and assist with auto completing forms. see par 100 - Example embodiments make of intent-based ML to facilitate software integrations and hyperautomation. An intent represents the purpose of a user's input. Example embodiments include a platform to map an intent for each type of user request to the API and applications supported. Coupling entity extraction in the natural language processing (NLP) of the invention enables a term or object that is relevant to your intents and that provides a specific context for an intent. In the platform according to example embodiments, example intents include adding connected systems and building or modifying integrations. see par 114 - In example embodiments, smart integration may include automatically performed tasks, or tasks performed based on a given environment such as, e.g., automatically updating workflows and automatically updating various services such as email and the like.);
automatically identifying, by the autonomous agent, a target task of the plurality of tasks to be executed (Applicant’s [0042] as published states “At step 320, the method can identify a target process including a selected task that matches the user query. The target process is one of a plurality of processes included in a process workflow. The target process includes a plurality of tasks and links connecting the plurality of tasks. The links define an operation sequence of the plurality of tasks.” [0090] as published states “Activities represent work that needs to be performed within the process. These can be either tasks, which are atomic activities, or sub-processes, which are activities composed of smaller tasks. Activities can be further classified based on their nature—manual tasks (requiring user intervention) and automated tasks (system tasks that are executed without user intervention). For instance, a user task might require an employee to approve a purchase order, while a system task might automatically generate a purchase requisition based on predefined criteria.
Touati see par 96 - With the answers to these questions, example embodiments build numerous integrations, robotic process automations, and business workflows. An example of this would be an integration to move incidents from ServiceNow to Jira. A user may say “I would like to open a bug in Jira for each urgent incident in ServiceNow in real time.” By doing so, a conversational agent may be configured to extract the source, target, event, action, and frequency of the integration, and set it up for the end user. see par 113 - The platform according to example embodiments can also serve to provide the following functionality: i) Automated testing of workflows and end-point confirmation analysis; see par 114 - In example embodiments, smart integration may include automatically performed tasks, or tasks performed based on a given environment such as, e.g., automatically updating workflows and automatically updating various services such as email and the like;
see also Lu – page 6, 2nd paragraph - The invention claims a method for flexibly configuring task flow of government administration system based on AI large model, decoupling the task processing flow and the core logic of the system, allowing the service personnel to flexibly configure and modify the task flow under the condition without programming ability, making the system more intelligent; see page 9, 1st paragraph – the embedded vector can be generated according to the service description content, and then the generated embedded vector is performed similarity search in the vector database so as to determine at least one associated vector and the corresponding task number.);
automatically prompting a … language model, by the autonomous agent, with a set of tools bound to the process workflow, the user query and the context prompt (Applicant’s [0026] as published states “During the design phase, the administrator 104 can also utilize a tools adaptor 138 of the workflow assistant 130 to generate a set of tools 148, which specify one or more application programming interfaces (APIs) used to perform tasks involved in a specific workflow. In some examples, the tools adaptor 138 can be configured to parse a document containing specifications of the APIs provided by an API provider 128 and represent each tool 148 as an API object. [0027] as published states “ After the set of tools 148 is generated, the administrator 104 can bind these tools to a specific workflow. This binding allows the APIs specified by the tools 148 to be used to implement the tasks involved in the workflow. For each tool, the administrator 104 can define a class or code for invoking the corresponding API and receiving the return results from the API call. Different sets of tools 148 can be mapped to different workflows, providing flexibility and customization based on the requirements of each workflow.
Touati see par 52 - to add additional capabilities of the method, a platform configured to implement the method may provide an API and can be integrated with tools that facilitate enterprise integration; see par 100 - An intent represents the purpose of a user's input. Example embodiments include a platform to map an intent for each type of user request to the API and applications supported. Coupling entity extraction in the natural language processing (NLP) of the invention enables a term or object that is relevant to your intents and that provides a specific context for an intent. In the platform according to example embodiments, example intents include adding connected systems and building or modifying integrations; see par 109 - integrating with external chatbots: Integrated systems may contain chatbots or be chatbot systems, in which case the platform can redirect users to specific chatbots when appropriate based on the user's context;
Lu, see page 5, 3rd paragraph - API (Application Programming Interface): It is a set of rules and tools for communication between different
software applications.) to determine whether execution of the target task requires invocation of any application programming interface, wherein the set of tools specify one or more application programming interfaces used to perform tasks involved in the process workflow (Touati – see par 60 - Example embodiments can also allow for the discovery, also referred to as self-discovery, of APIs either through ingestion of standardized API catalogs or through a self-discovery process of making use of test-cases and examining inputs/outputs and validating the API and its subsequent integration. Example embodiments provide the opportunity to discover and integrate multiple APIs from a comprehensive range of enterprise business software platforms. see par 101 - Example embodiments can also enable mapping of intent to entities that may be pre-defined in the platform either by manual coding or by automated discovery or application programming interface (API) discovery process that can automatically map the functionality of the API along with the possible functional manipulations which are possible with given entities; see par 102 - Example embodiments create an agnostic speech interface for inclusion of different third-party, and/or more specialized bots, as well as integration of different code or low-code systems. Example embodiments provide added flexibility due to the use of speech to interrogate the actual end-user data itself and its ability to work with a wide range of application APIs that can be auto-discovered and exposed to the speech/bot interface.
Lu page 8, 7th paragraph - the invention realizes the step planning of the government system task by the mode of prompting the engineering and mounting the government knowledge base, and outputs the API scheduling code of each step by using the large model code generating ability, and finally realizes the task scheduling by executing the code. see page 9, 1st paragraph - Further, in the relational database, the corresponding target solution can be determined according to the task number, and then the target solution is input into a preset large model as prompt information to generate a target code for invoking each API resource and finishing the related task requirement.);
Touati discloses chatbots and “natural language processing” (See par 98-102; (See par 59 - NLP is coupled with Chatbot technology to enable text and/or voice based interaction with the system; In example embodiments, a user may be able to communicate to the both in natural language and/or low- or no-code development environments as well as to ask status of the integration involving a specific component of the enterprise software or other queries of the specific integration or the global systemic ramifications thereof).
Lu discloses:
automatically prompting a “large language model,” by the autonomous agent, with a set of tools bound to the process workflow, the user query, and the context prompt to determine whether execution of the target task requires invocation of any application programming interface, wherein the set of tools specify one or more application programming interfaces used to perform tasks involved in the process workflow (Lu – see page 10, 2nd -3rd paragraphs - step S101, generating an embedded vector according to the service description content; step S102, performing similarity search to the embedded vector based on the preset vector database, determining at least one associated vector and corresponding task number; see page 10, 2nd to last paragraph - the government system in the embodiment of the invention can generate embedded vector by the service description content input by the user (disclosing “user query”), in the vector database, performing similarity search with the existing corpus vector, obtaining topK vectors and corresponding task ID with similar semantics, in the relational database, according to the text content corresponding to the ID retrieval task and the calling solution, further inputting K solutions in the database as prompt (disclosing context prompt) to the ChatGLM large model, and prescribing the format of the output solution, so as to generate a target code, It is used for calling each API resource and finishing the related task requirement; see page 14, 3rd paragraph - Large Language Model (LLM): The invention claims an artificial intelligent model based on deep learning, which can be used for understanding and generating tasks aiming at various natural languages and supports the use of industry data for fine adjustment so as to adapt to different business scenes; Open-source large model comprises GPT-3, ChatGLM).
Touati and Lu disclose:
receiving, by the autonomous agent, a response generated (Touati see par 109 - integrating with external chatbots: Integrated systems may contain chatbots or be chatbot systems, in which case the platform can redirect users to specific chatbots when appropriate based on the user's context) by the large language model (Lu discloses entire limitation– see page 10, 2nd to last paragraph - the government system in the embodiment of the invention can generate embedded vector by the service description content input by the user (disclosing “user query”), in the vector database, performing similarity search with the existing corpus vector, obtaining topK vectors and corresponding task ID with similar semantics, in the relational database, according to the text content corresponding to the ID retrieval task and the calling solution, further inputting K solutions in the database as prompt (disclosing context prompt) to the ChatGLM large model, and prescribing the format of the output solution, so as to generate a target code (disclosing “response generated by large language model”), It is used for calling each API resource and finishing the related task requirement; The invention uses the vector similarity search to realize the prompt construction, optimizes the large model token number limit problem, provides the solution for obtaining the optimal generated content in the limited input length, and improves the controllability of the large model; see page 10, last paragraph - generates an embedded vector according to the service description content through the above solution; performing similarity search on the embedded vector based on a preset vector database, determining at least one associated vector and a corresponding task number; determining at least one target solution in the preset relational database according to the task number); and
for the response indicating a specific application programming interface for executing the target task, automatically determining, by the autonomous agent, whether user intervention is required before invoking the specific application programming interface (Touati – see par 52 - to add additional capabilities of the method, a platform configured to implement the method may provide an API and can be integrated with tools that facilitate enterprise integration (such as Salesforce, IBCO, Boomi); see par 61 - API discovery can be further advanced with the inclusion of mapping intent to natural language commands.
see par 101 - Example intents include adding connected systems and building or modifying integrations. Each intent may have a desired or predetermined parameters associated therewith, which the bot attempts to identify by prompting the user(s) for the specific values of each parameter (slot filling). When possible, machine learning is used during slot filling to avoid a user having to memorize specific terms or vocabulary. If the parameters are expressed in the initial intent, the bot may be configured to directly extract them and avoid the need for additional prompts;
Lu – see page 13, #6-8 - The large model generation scheme is returned to the user for confirmation, and the API is selected according to the scheme after the confirmation is correct; 7, API selector according to large model each step output content, combining vector database professional knowledge, from service resource pool selected from the most suitable API, to form API queue; 8, code generating model according to API queue and context, generating executable code and filling API parameter);
responsive to determining that user intervention is required, dynamically adjusting, by the autonomous agent, the process workflow to incorporate an input received via the user interface before invoking the specific application programming interface (Touati – see par 65 - The platform according to example embodiments also provides an ‘escalation’ path to make use of natural language questions to the virtual developer, which may reach out to the integrated systems to attempt to formulate an answer. If it needs more specifics, it may prompt the user for them before responding. see par 101 - Example embodiments can also enable mapping of intent to entities that may be pre-defined in the platform either by manual coding or by automated discovery or application programming interface (API) discovery process. Example intents include adding connected systems and building or modifying integrations. Each intent may have a desired or predetermined parameters associated therewith, which the bot attempts to identify by prompting the user(s) for the specific values of each parameter (slot filling).
Lu – see page 12, 1st paragraph - the K target solutions in the database are input to the ChatGLM large model as prompts, and the format of the output solution is specified, the large model generates the solution content and performs confirmation, wherein the solution content can be returned to the user for confirmation, and/or The API is selected according to the solution after confirming that there is no error through the model automatic identification confirmation);
responsive to determining that user intervention is not required, automatically invoking, by the autonomous agent, the specific application programming interface to execute the target task (Touati – see par 60 - Example embodiments can also allow for the discovery, also referred to as self-discovery, of APIs either through ingestion of standardized API catalogs or through a self-discovery process of making use of test-cases and examining inputs/outputs and validating the API and its subsequent integration. see par 61 -API discovery can be further advanced with the inclusion of mapping intent to natural language commands. In this way, the system can expand in compatibility with enterprise software systems with or without human integration efforts either through direct software engineering between the enterprise systems or engineering of the platform according to example embodiments that affords such integration.
see also Lu – see page 8, 2nd to last paragraph - the task flow auxiliary arrangement can be performed through interaction with AI implementation, after the large model identifies the user intention, the approximate flow stored in the system is inquired through the knowledge map, the task execution API is generated to invoke the node order, according to the user intention, the parameter is automatically filled, the node is arranged for the second time and modified in sequence through the multi-round conversation or the manual debugging of the user page terminal, and the node is stored in the database after confirming and then the flow is issued; see page 9, 7th paragraph - the most suitable API can be selected from the service resource pool according to the solution content output by each step of the preset large model, in combination with the professional knowledge of the vector database, to form an API queue, the code generation model according to the API queue and the context, generating the executable code and filling the API parameter so as to execute the execution code to obtain the execution result.); and
generating an output on the user interface based on the response (Touati – see par 110 - The system may be configured to expose discovered APIs as chatbots. For any given API, there is a set of required and optional fields. When combined with the context of the purpose of each API, the platform may be configured to generate a unique chat experience for each connected API;
Lu – see page 13, #6-8 - The large model generation scheme is returned to the user for confirmation, and the API is selected according to the scheme after the confirmation is correct; 7, API selector according to large model each step output content, combining vector database professional knowledge, from service resource pool selected from the most suitable API, to form API queue; 8, code generating model according to API queue and context, generating executable code and filling API parameter).
Both Touati and Lu are analogous art as they are directed to forming process flow/workflows for business/ERP processes while analyzing natural language input from users (See Touati Abstract, par 53, 60, 80; Lu Abstract, page 14 (“different business scenes”)). Touati discloses chatbots and “natural language processing” (See par 59, 98-102). Lu improves upon Touati by disclosing large language models used for analyzing user natural language input/queries. One of ordinary skill in the art would be motivated to further include having large language models to efficiently improve upon the design of integrations using bots (See par 53) and for ERP systems (See par 80) in Touati.
Accordingly, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the system and method of integrating services between different software systems (Abstract) in Touati to further use large language models as disclosed in Lu, since the claimed invention is merely a combination of old elements, and in combination each element merely would have performed the same function as it did separately, and one of ordinary skill in the art would have recognized that the results of the combination were predictable and there is a reasonable expectation of success.
Concerning independent claim 11, Touati discloses:
A computer-implemented method for improving process flow automation in an enterprise resource planning (ERP) platform (Touati – see par 54 - creation of an integration or process automation via a chatbot is illustrated in FIGS. 5A-5D; see par 80 -example of an integration flow, user 102 may build an integration scenario that synchronizes data between an enterprise resource planning (ERP) system and a customer relationship management (CRM) system; see par 165 - As shown in FIG. 15, computer system/server 12 in cloud computing node 10 is shown in the form of a general-purpose computing device. The components of computer system/server 12 may include, but are not limited to, one or more processors or processing units 16), the method comprising.
The remaining limitations are similar to claim 1. Claim 11 is rejected for the same reasons as claim 1.
It would have been obvious to combine Touati and Lu for the same reasons as discussed with regards to claim 1.
Concerning independent claim 20, Touati and Lu disclose:
One or more non-transitory computer-readable media having encoded thereon computer-executable instructions (Touati – see par 34-35 - The computer readable storage medium can be a tangible device that can retain and store instructions for use by an instruction execution device; see par 168-170 – system memory 28 includes system readable media; program/modules 42 stored in memory carry out functions/embodiments) causing one or more processors to perform a method for improving process flow automation in an enterprise resource planning (ERP) platform, the method comprising (Touati – see par 54 - creation of an integration or process automation via a chatbot is illustrated in FIGS. 5A-5D; see par 80 -example of an integration flow, user 102 may build an integration scenario that synchronizes data between an enterprise resource planning (ERP) system and a customer relationship management (CRM) system; see par 165 - As shown in FIG. 15, computer system/server 12 in cloud computing node 10 is shown in the form of a general-purpose computing device. The components of computer system/server 12 may include, but are not limited to, one or more processors or processing units 16), the method comprising.
The remaining limitations are similar to claim 1. Claim 11 is rejected for the same reasons as claim 1.
It would have been obvious to combine Touati and Lu for the same reasons as discussed with regards to claim 1.
Concerning claims 2 and 12, Touati discloses using semantic dictionary to map words to other words having similar meaning to link field names between assets and make connections between fields and assets (See par 89) and having conversational querying of content sources via automated interactive conversational environment (See par 129).
Lu discloses:
The computing system of claim 1, wherein the operations further comprise embedding the user query into a query vector (Lu – see page 10, 2nd paragraph - step S101, generating an embedded vector according to the service description content).
It would have been obvious to combine Touati and Lu for the same reasons as discussed with regards to claim 1. Lu improves upon Touati by disclosing having a “vector” for service description.
Concerning claims 3 and 13, Touati and Lu disclose:
The computing system of claim 2, wherein identifying the target process comprises measuring similarities between the query vector and a plurality of process vectors representing the plurality of processes included in the process workflow (Lu – see page 10, 3rd paragraph - step S102, performing similarity search to the embedded vector based on the preset vector database, determining at least one associated vector and corresponding task number; see page 10, 2nd to last paragraph - the government system in the embodiment of the invention can generate embedded vector by the service description content input by the user, in the vector database, performing similarity search with the existing corpus vector, obtaining topK vectors and corresponding task ID with similar semantics, in the relational database, according to the text content corresponding to the ID retrieval task and the calling solution, further inputting K solutions in the database as prompt to the ChatGLM large model, and prescribing the format of the output solution, so as to generate a target code, It is used for calling each API resource and finishing the related task requirement.).
It would have been obvious to combine Touati and Lu for the same reasons as discussed with regards to claim 1 and 2. Touati discloses using semantic dictionary to map words to other words having similar meaning to link field names between assets and make connections between fields and assets (See par 89) and having conversational querying of content sources via automated interactive conversational environment (See par 129). Lu improves upon Touati by disclosing having a “vector” for service description, and then using the vectors to look for similarities.
Concerning claims 4 and 14, Touati and Lu disclose:
The computing system of claim 3, wherein the operations further comprise generating the plurality of process vectors by embedding respective context prompts describing the plurality of processes included in the process workflow (Lu – see page 10, 2nd to last paragraph - The invention uses the vector similarity search to realize the prompt construction, optimizes the large model token number limit problem, provides the solution for obtaining the optimal generated content in the limited input length, and improves the controllability of the large model; see page 10, last paragraph - generates an embedded vector according to the service description content through the above solution; performing similarity search on the embedded vector based on a preset vector database, determining at least one associated vector and a corresponding task number; determining at least one target solution in the preset relational database according to the task number).
It would have been obvious to combine Touati and Lu for the same reasons as discussed with regards to claim 1-3.
Concerning claims 5 and 15, Touati and Lu disclose:
The computing system of claim 4, wherein the operations further comprise generating the context prompts describing the plurality of processes included in the process workflow, wherein generating the context prompt for a selected process comprises prompting the large language model with a prompt including an object representing the selected process (Applicant’s [0023] as filed gives example “process object can be represented in a data exchange format, such as JavaScript Object Notation (JSON) or the like”; [0043] as filed gives example “parse a document containing specifications of the one or more APIs and represent each tool as an API object containing information of a corresponding API”
Touati – see par 100 - Example embodiments make of intent-based ML to facilitate software integrations and hyperautomation. An intent represents the purpose of a user's input. Example embodiments include a platform to map an intent for each type of user request to the API and applications supported. Coupling entity extraction in the natural language processing (NLP) of the invention enables a term or object that is relevant to your intents and that provides a specific context for an intent. In the platform according to example embodiments, example intents include adding connected systems and building or modifying integrations.
Lu discloses the limitations based on broadest reasonable interpretation in light of the specification – including the “large language model” –see page 9, 5th paragraph - large language basic model can provide strong dialogue on the open domain task, The context learning and code generation capability can also generate a high-level solution outline for the specific field task; see page 9, 7th paragraph - the most suitable API can be selected from the service resource pool according to the solution content output by each step of the preset large model, in combination with the professional knowledge of the vector database, to form an API queue, the code generation model according to the API queue and the context context, generating the executable code and filling the API parameter so as to execute the execution code to obtain the execution result; see page 13, #7-8 - API selector according to large model each step output content, combining vector database professional knowledge, from service resource pool selected from the most suitable API, to form API queue; 8: code generating model according to API queue and context context, generating executable code and filling API parameter).
It would have been obvious to combine Touati and Lu for the same reasons as discussed with regards to claim 1-3.
Concerning claims 6 and 16, Touati discloses “mapping can be further informed and refined by both manual input and refinements to machine learning systems coupled with rules-based algorithms that define software specific systems as well as general logic that is generically applicable to a wider range of software systems”, discloses integrating files to allow documents to flow from ServiceNow to another Service (See par 54). Touati discloses “a client component of web channel 130 may include a hypertext markup language (HTML) file with JS and CSS files included in a website.”
Lu discloses:
The computing system of claim 5, wherein the operations further comprise parsing a markup language definition of the process workflow, and representing the selected process as a set of nodes in the object, wherein the set of nodes represent the tasks of the selected process and are organized in a hierarchical relationship representing the operation sequence of the tasks (Applicant has example in [0022] of “markup language like XML”
Lu discloses – see page 12, 9th paragraph - after the flow is executed successfully, the flow can be issued and stored in the database. The
published flow is stored in the system structured database, and the mapping relationship between the flow and the XML file is established through the field, which is convenient for the quick calling in the subsequent business flow arranging process; see page 14, last paragraph - XML file: An extensible markup language is used for storing and transmitting structured data. A tag is generally used in a file to mark data elements and to represent relationships and hierarchical structures between data by nesting and attributes. The method stores the flow content in the XML file, and defines the interface ID, interaction type (manual, automatic synchronization, automatic synchronization), executor, whether to start, node type (handling, handling) for each node. XML files record flow nodes in sequence, layering is realized by nesting, the outermost layer uses process label to represent the entire flow, each flow needs to include the beginning, end node, respectively, using start and end label, label comprises display name (displayname), shape (shape), ID, type (type) and so on. the task node is represented by task, which is also embedded in the process, and is in parallel relationship with the starting and ending nodes, displayed according to the flow sequence;
It would have been obvious to combine Touati and Lu for the same reasons as discussed with regards to claim 1 and 2. Touati discloses “mapping can be further informed and refined by both manual input and refinements to machine learning systems coupled with rules-based algorithms that define software specific systems as well as general logic that is generically applicable to a wider range of software systems”, discloses integrating files to allow documents to flow from ServiceNow to another Service (See par 54). Touati discloses “a client component of web channel 130 may include a hypertext markup language (HTML) file with JS and CSS files included in a website.” Lu improves upon Touati by disclosing using XML to define hierarchical structures.
Concerning claim 7, Touati and Lu disclose:
The computing system of claim 1, wherein the user query is one of a plurality of user queries received in a query session, wherein prompting the large language model includes sending a history of query session to the large language model, wherein the history of the query session stores the plurality of user queries and corresponding responses generated by the large language model (Touati – see par 65 - There is also the ability to examine integration history so as to inform the system if the integration is a “re-do” of a past integration versus a novel integration. There is also the ability to provide personalized NLP and intent/entity optimization based on individual ML profile which is optimized based on the individuals use, metadata and integration performance history. see par 107 - Example embodiments may learn from the selected option selections as well as feedback from the system (including violation notices, analytics, measures of data integrity) to better recommend these suggestions to individual users and organizations.
Lu discloses “large language model”–see page 9, 2nd to last paragraph - after the execution result is fed back to the target user, the satisfaction evaluation of the user can be collected for optimizing the preset large model; The execution result of this time can be used as the data for strengthening learning, and the planning ability of the model solution can be continuously improved; see page 11, 6th paragraph - obtaining historical question and answer text data; making a corresponding task flow according to the historical question and answer text data to form a government affair service data set; training based on the government affair service data set to obtain the preset large model; page 11, 7th paragraph - by collecting the administrative language material data set, namely the historical question and answer text data, and by means of fine adjustment, the expression and reasoning planning ability of the large model under the administrative scene is enhanced, so that it has higher correctness and reliability when processing the dialogue and task flow of the administrative scene, It provides support for the government system.)
It would have been obvious to combine Touati and Lu for the same reasons as discussed with regards to claim 1.
Concerning claims 8 and 17, Touati and Lu disclose:
The computing system of claim 1, automatically identifying the target task is based on the operation sequence (Touati – see par 96 - With the answers to these questions, example embodiments build numerous integrations, robotic process automations, and business workflows. An example of this would be an integration to move incidents from ServiceNow to Jira. A user may say “I would like to open a bug in Jira for each urgent incident in ServiceNow in real time.” By doing so, a conversational agent may be configured to extract the source, target, event, action, and frequency of the integration, and set it up for the end user.
Lu – see page 8, 6th paragraph - the invention realizes the step planning of the government system task by the mode of prompting the engineering and mounting the government knowledge base, and outputs the API scheduling code of each step by using the large model code generating ability, and finally realizes the task scheduling by executing the code. 7th paragraph - at the page end of the system worker, the task flow auxiliary arrangement can be performed through interaction with AI implementation, after the large model identifies the user intention, the approximate flow stored in the system is inquired through the knowledge map, the task execution API is generated to invoke the node order, according to the user intention, the parameter is automatically filled, see page 12, 9th paragraph - after the flow is executed successfully, the flow can be issued and stored in the database. The published flow is stored in the system structured database, and the mapping relationship between the flow and the XML file is established through the field, which is convenient for the quick calling in the subsequent business flow arranging process).
It would have been obvious to combine Touati and Lu for the same reasons as discussed with regards to claim 1.
Concerning claims 9 and 18, Touati and Lu disclose:
The computing system of claim 1, wherein the response includes related metadata required to invoke the application programming interface (Applicant’s [0032] as published - related metadata (e.g., required API parameters, API endpoint, etc.).
Touati - see par 110 - For any given API, there is a set of required and optional fields.
Lu – see page 9, 1st paragraph; page 10, 9th paragraph - further inputting K solutions in the database as prompt to the ChatGLM large model, and prescribing the format of the output solution, so as to generate a target code, It is used for calling each API resource and finishing the related task requirement; see page 13, #6-8 - The large model generation scheme is returned to the user for confirmation, and the API is selected according to the scheme after the confirmation is correct; 7, API selector according to large model each step output content, combining vector database professional knowledge, from service resource pool selected from the most suitable API, to form API queue; 8, code generating model according to API queue and context, generating executable code and filling API parameter).
It would have been obvious to combine Touati and Lu for the same reasons as discussed with regards to claim 1.
Concerning claims 10 and 19, Touati and Lu disclose:
The computing system of claim 9, wherein the operations further comprise: responsive to determining that user intervention is required, prompting a user input on the user interface, and conditioning invocation of the application programming interface based on the user input (Touati – see par 65 - The platform according to example embodiments also provides an ‘escalation’ path to make use of natural language questions to the virtual developer, which may reach out to the integrated systems to attempt to formulate an answer. If it needs more specifics, it may prompt the user for them before responding.
Lu – see page 12, 1st paragraph - the K target solutions in the database are input to the ChatGLM large model as prompts, and the format of the output solution is specified, the large model generates the solution content and performs confirmation, wherein the solution content can be returned to the user for confirmation, and/or The API is selected according to the solution after confirming that there is no error through the model automatic identification confirmation).
It would have been obvious to combine Touati and Lu for the same reasons as discussed with regards to claim 1.
Response to Arguments
Applicant's arguments filed 5/11/26 have been fully considered but they are not persuasive and/or are moot in view of the new rejections.
Applicant argues new limitations for claim 1 not being in Touati. Remarks, pages 9-11. In response, Examiner respectfully disagrees. The arguments are moot in view of the revised rejections with new citations.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure:
Vidgof, “Large Language Models for Business Process Management: Opportunities and Challenges,” 2023, In International Conference on Business Process Management Forum, Springer Nature Switzerland, pages 107-122 – directed to having LLM in process discovery for business processes (see Abstract, page 114)
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to IVAN R GOLDBERG whose telephone number is (571)270-7949. The examiner can normally be reached 830AM - 430PM.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Anita Coupe can be reached at 571-270-3614. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/IVAN R GOLDBERG/Primary Examiner, Art Unit 3619
Prosecution Insights