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 .
Response to Arguments/Amendment
Applicant's arguments/amendments filed July 21, 2026 have been fully considered.
Applicant’s arguments with respect to claim(s) 1 have been considered but are moot because the new ground of rejection does not rely on any reference applied in the prior rejection of record for any teaching or matter specifically challenged in the argument.
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 (i.e., changing from AIA to pre-AIA ) 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 text of those sections of Title 35, U.S. Code not included in this action can be found in a prior Office action.
Claims 1-29 are rejected under 35 U.S.C. 103 as being unpatentable over Vasquez et al. (US 12130821 B1, hereinafter “Vasquez”) in view of Doshi et al. (US 20240264874 A1, hereinafter, “Doshi”) and Payne et al. (GB 2610902 A, hereinafter Payne).
Regarding claim 29,
Vasquez teaches:
An apparatus for orchestrating model operations for digital asset processing (Vazquez: Fig. 2-3, col 17 lines 50-60, “DCAMPS-C system”),
comprising:
a model operations orchestrator configured to function as a communication layer between a plurality of containerized environments, wherein each containerized environment houses one or more modules configured to perform specific tasks on a digital asset (Vasquez: col 2 lines 40-47, “. . . the DCAMPS-C system further comprises a DCAMPS orchestrator that is configured to manage deployment of the DCAMPS-C container and scaling of the components encapsulated in the deployed DCAMPS-C container. . .”; col. 17: lines 50-61, “. . . a DCAMPS-C container (at 210) . . . a container orchestrator (at 240), a data classification engine (at 245), an automated metadata publishing service (at 250), and an LLM container (at 255)”; col 19: lines 5-11, “. . . the LLM container (at 255) is a core component of the YOLAGPT LLM module (at 215). Specifically, the LLM container (at 255) includes the YOLAGPT large language model used to provide large language model capabilities for the DCAMPS-C system, by way of the YOLAGPT LLM module (at 215). . .”; col 17 lines 29-39, “. . . the generative pre-trained transformer LLM module is an optimized learning, artificial-intelligence, generative pre-trained transformer LLM module. . . the YOLAGPT LLM module is configured to (i) interpret natural language input, (ii) translate the interpreted natural language input into formal rules, (iii) integrate the translated formal rules into the data classification engine, and (iv) scan and classify data assets and content across the plurality of data sources of the organization based on the classification rules. . .”; col 1 lines 64-67; “. . . a data classification engine capable of accurately classifying and categorizing data. . .; col 2 lines 5-10, “. . .an automated metadata publishing service that is configured to receive the classified data and generated metadata from the data classification engine and perform a metadata enrichment process to add the metadata to the digital assets and content of the organization. . .”; NOTE: The plurality of containerized environment is the DCAMPS-C container 210, and the LLM container. The container orchestrator functions as a communication layer between the containers as illustrated in Fig. 2. The LLM container houses an LLM module with a specific tasks acting as a generative pre-trained transformer module. The DCAMPS container houses data classification engine and automated metadata publishing service 250 modules which perform specific tasks on digital assets as referenced above.);
an input interface configured to receive a definition of an orchestration pipeline, the orchestration pipeline specifying a sequence of tasks to be performed on the digital asset, each task corresponding to a particular module within a particular containerized environment (Vasquez: col 18 lines 25-35, “. . . the user (at 205) . . . interacts with (at 260) . . . the user (at 205) may provide natural language input to the DCAMPS-C container (at 210). . .”; col: 20, “lines 64-67, “. . . input devices enable an end user to communicate information. . . the user may provide natural language input. . .”; col 15 lines 40-51, “. . . the YOLAGPT LLM module is the component which performs the step for interpreting the natural language input and translating the interpreted natural language input into formal rules (at 140) . . . Next, the DCAMPS-C and YOLAGPT LLM module deployment, configuration, and runtime process 100 proceeds to a step for scanning and classifying data sources based on defined rules (at 145). Specifically, the step for scanning and classifying (at 145) is performed by the YOLAGPT LLM module to the data sources based on the classification rules (defined at 130). . .”; NOTE: The input interface is the interface at 260 where the user inputs a natural language. This language is then translated into formal rules defining the orchestration pipeline configuration based on the natural language. Therefore, the formal rules translated from the language input from an input interface defines the orchestration pipeline configuration, deployment, and runtime process. Vasquez orchestration pipeline sequence of tasks to be performed on the digital asset is shown in Fig .1. As discussed above, the LLM container and DCAMPS container have particular modules within them that correspond to each task. Also see Fig. 2.);
a dynamic loader configured to dynamically load one or more pipeline modules based on the orchestration pipeline, wherein each pipeline module exposes an interface that allows the model operations orchestrator to provide specific configuration parameters required for execution of a corresponding task for that pipeline module (Vasquez: col 11 lines 65-67 to col 12 lines 1- 51, “. . . the components of DCAMPS-C work together to create a comprehensive solution that automates data classification and metadata generation processes, streamlines deployment and scaling, and enhances overall efficiency in managing digital assets and content. The containerized nature of the system ensures consistency and seamless integration with modern IT infrastructures . . . steps and sub-steps are driven by an ordered process that may perform all steps in sequence or in a different order based on (in part) a programming paradigm or need-based order. For instance, many sub-routines may be functions in object classes called by a main program to execute their respective tasks and return the results (data). The functions may inevitably involve one or more decision-making algorithms which are configured to analyze the content and metadata of data sources to determine the appropriate classification based on customizable rules and policies. Furthermore, such decision-making algorithms may use machine learning or artificial intelligence (AI) techniques to improve classification accuracy over time. Metadata generation also involves a programmatic-like logic that may include the use of AI, machine learning, or pre-defined rules to analyze the content and generate relevant metadata for digital assets, thereby improving searchability and discoverability. Data source integration may have logic processing steps that involve connecting to various data sources, such as file servers, databases, SharePoint, and web sources, and retrieving data for classification and metadata generation. In order to carry out such processing steps, the data source integration may utilize one or more application programming interface(s) (APIs), authentication protocols, and/or data access methods specific to each data source. The containerization logic includes the processes and configurations required to package the software and its dependencies within a container, ensuring a consistent and isolated environment for running the software across different platforms and deployment environments. Finally, the DevOps and cloud-native infrastructure integration logic may be based on an implementation of best practices or may involve use of tools and/or workflows that enable seamless deployment, scaling, and management of DCAMPS-C in modern IT environments. . .”; NOTE: The DCAMPS-C system is functions as a dynamic loader that loads pipeline modules such as the LLM module in the LLM container and the classification and publishing modules in the DCAMPS container. As discussed above, the formal rules that define the configuration of the orchestration pipeline to automate data classification and metadata generation processes across different platforms. Therefore the system dynamically sets-up the configuration depending on the type of platform and deployment environments to support automation. The exposed interface includes API for data integration that is specific to each data source making the pipeline acquire specific configuration parameters corresponding to each data source.);
a scheduler configured to schedule execution of the one or more pipeline modules in accordance with the orchestration pipeline, wherein the one or more pipeline modules are executed either sequentially or in parallel based on the sequence of tasks (NOTE: See Vasquez col 11 lines 65-67 to col 12 lines 1- 51. Vasquez DCAMPS-C functions as a scheduler because the pipeline configuration “steps and sub-steps are driven by an ordered process that may perform all steps in sequence or in a different order”. Therefore the execution of the pipeline modules with specific tasks/steps in accordance with the orchestration pipeline is executed sequentially based on the sequence of tasks and scheduled in an ordered process.)
and a processing engine configured to process the digital asset through the sequence of tasks as defined in the orchestration pipeline (Vasquez: col 1 lines 62-67 to col 2 lines 1-3, “. . . the DCAMPS-C system comprises (i) a data classification engine capable of accurately classifying and categorizing data based on classification rules and policies that provide flexibility and accuracy in data classification for data of an organization, wherein the data classification engine is further configured to classify the data (“classified data”) and generate metadata for the classified data. . .”; NOTE: The data classification engine is the processing engine that processes digital assets according to the rules configured for the orchestration. The digital assets are processed through the sequence of tasks as defined in the orchestration pipeline as set by the formal rules from scanning – classification – publish.).
Although Vasquez system schedules the execution of pipeline modules in sequence or driven by an ordered process, Vasquez does not explicitly disclose that the processing is executed without collision.
The analogous art Doshi teaches:
processing digital assets in an orchestration pipeline without collision (Doshi: ¶168-171, “A barrier mechanism may be used to synchronize workloads. For instance, a barrier may pause, suspend, sleep, or put a workload in a wait state to allow it to synchronize with one or more other workloads that are going to rendezvous at the barrier . . . interoperability architecture for transferring or exchanging digital assets. . .”; NOTE: Vasquez (claim 8) and Doshi (¶197) both comprise a Kubernetes container for the orchestration pipeline, which is also an alternative disclosed in the applicant specification page 21 lines 19-25. Doshi’s barrier mechanism allows for processing digital assets without collision by preventing race conditions where multiple functions trying to use resources without waiting for other functions to finish processing first, leading to data corruption. With Doshi’s barrier mechanism, a workload is put to a wait state and allow proper synchronization with other workloads.).
It would have been obvious to a person having ordinary skill in the art (PHOSITA) before the effective filing date of the claimed invention to combine Vasquez, Doshi, and Payne and include processing digital assets in an orchestration pipeline without collision by implementing Doshi’s barrier mechanism.
The reason for doing so to “pause, suspend, sleep, or put a workload in a wait state to allow it to synchronize with one or more other workloads that are going to rendezvous at the barrier” (Doshi: ¶168).
Although Vasquez system processes the digital asset through a sequence of tasks in the orchestration pipeline, Vazquez does not disclose wherein the digital asset is managed by the model operations orchestrator as a single manifest asset that is handled throughout the orchestration pipeline.
The analogous art Doshi teaches:
wherein the digital asset is managed by the model operations orchestrator as a single manifest asset that is handled throughout the orchestration pipeline (Doshi: ¶108, “. . . The top-half update package 1504 may include a Bill of Materials (BOM) describing software, firmware, drivers, or hardware components. The BoM may include attestation reference values for matching attestation evidence, endorsed values that represent attributes and properties of the component known to the component vendor, supplier, or deployment agent. The bottom-half update package 1514 may contain updated firmware, software, settings, configuration parameters, calibration parameters, policies or other information related to applying the update. . . .”¶122, “Manifest standards contain BOM lifecycle intents where updates and patches are tracked by a BoM.” NOTE: The BOM acts as a single manifest asset that tracks update to the digital assets when processed in the orchestration pipeline.)
It would have been obvious to a person having ordinary skill in the art (PHOSITA) before the effective filing date of the claimed invention to combine Vasquez, Doshi, and Payne and include: wherein the digital asset is managed by the model operations orchestrator as a single manifest asset that is handled throughout the orchestration pipeline.
The reason for doing so is to maintain “a database of update dependencies such that older images that are not aware of newer images can be searched efficiently” (Doshi: ¶122).
Although Vasquez teaches: performing specific tasks on digital assets, Vasquez, fails to teach: perform specific transformation tasks on a visual digital asset, the specific transformation tasks comprising at least one of converting, optimizing, or rendering the visual digital asset.
And although Vasquez teaches: an input interface configured to receive a definition of an orchestration pipeline, the orchestration pipeline specifying a sequence of tasks to be performed on the digital asset, Vasquez fails to teach specific transformation task to be performed on a visual digital asset comprising at least one of converting, optimizing, or rendering the visual digital asset.
And although Vasquez teaches: each task corresponding to a particular module within a particular containerized environment; a dynamic loader configured to dynamically load one or more pipeline modules based on the orchestration pipeline, wherein each pipeline module exposes an interface that allows the model operations orchestrator to provide specific configuration parameters required for execution of a corresponding task for that pipeline module, Vasquez fails to teach specific transformation task to be performed on visual digital asset.
And although the combination of Vasquez, Doshi, and Payne teaches: a scheduler configured to schedule execution of the one or more pipeline modules in accordance with the orchestration pipeline, wherein the one or more pipeline modules are executed either sequentially or in parallel based on the sequence of tasks and without collision; and a processing engine configured to process the visual digital asset through the sequence of tasks as defined in the orchestration pipeline, wherein the digital asset is managed by the model operations orchestrator as a single manifest asset that is and handled throughout the orchestration pipeline, the combination of Vasquez, Doshi, and Payne fails to teach that the manifest is passed between the one or more pipeline modules.
(NOTE: Vasquez teaches the orchestration pipeline to process digital assets, however, Vasquez processes metadata as the digital assets, and not including transforming visual data asset to be optimized, converted or rendered. Doshi teaches the barrier synchronization method combined with Vasquez to prevent collision during in the execution of Vasquez orchestration pipeline. The combination of Vasquez, and Doshi fails to teach specific transformation tasks to be performed on visual digital asset, and the manifest is passed between the one or more pipeline modules.
The analogous art Payne teaches:
performing specific transformation tasks on a visual digital asset, the specific transformation tasks comprising at least one of converting, optimizing, or rendering the visual digital asset (Payne: ¶10, “. . . a 3D product model of a manufactured product for rendering as a digital asset, the method comprising: receiving product data in a structured form, the structure of the data defining different configuration fields of the product and a plurality of different instances of the product, each different instance of the product comprising a unique combination of values of the different configuration fields; creating a product type manifest describing a product type, the product type manifest specifying a set of types of aesthetic components which make up the product type; generating a product manifest describing an instance of the product from the plurality of different instances of the product, the instance of the product being of a product type defined in the product type manifest, the product manifest comprising a unique product identifier and a plurality of unique component identifiers of components making up the product, each of the unique component identifiers identifying a component belonging to the set of aesthetic components types defined by the product type manifest; retrieving a 3D component model for each component specified in the product manifest from a library of 3D component models; and assembling the retrieved 3D component models of the product manifest into a 3D product model of the product suitable for rendering to create a digital asset . . .”; ¶26, “. . . generating an (audio visual) digital asset comprising using a method of creating a 3D product model of an assembled product as described above and further comprising rendering the 3D product model to create the digital asset . . .”; NOTE: Payne’s pipeline: The product manifest describes an product >> generates product manifest including data and component identifiers, constituting to the claimed visual digital asset, for the product used for rendering >> assembling the visual digital assets after retrieval >> rendering the product model, constituting to rendering the visual digital assets.)
Payne further teaches:
a single manifest asset that is passed between the one or more pipeline modules (Payne: ¶10, “. . . 3D product model of a manufactured product for rendering as a digital asset, the method comprising: receiving product data in a structured form, the structure of the data defining different configuration fields of the product and a plurality of different instances of the product, each different instance of the product comprising a unique combination of values of the different configuration fields; creating a product type manifest describing a product type, the product type manifest specifying a set of types of aesthetic components which make up the product type; generating a product manifest describing an instance of the product from the plurality of different instances of the product, the instance of the product being of a product type defined in the product type manifest, the product manifest comprising a unique product identifier and a plurality of unique component identifiers of components making up the product, each of the unique component identifiers identifying a component belonging to the set of aesthetic components types defined by the product type manifest; retrieving a 3D component model for each component specified in the product manifest from a library of 3D component models; and assembling the retrieved 3D component models of the product manifest into a 3D product model of the product suitable for rendering to create a digital asset . . .”; NOTE: Payne’s pipeline/modules: generating product data in a structured form >> generate product TYPE manifest >> generate product manifest including data from the product type manifest, making the product manifest a single manifest asset >> [retrieving 3D data from the single manifest >> assembling the 3D component >> render]. The retrieval, assembly, and rendering are the one or more pipeline modules that the product manifest is passed in between.)
It would have been obvious to a person having ordinary skill in the art (PHOSITA) before the effective filing date of the claimed invention to combine Vasquez, Doshi, and Payne, and include: wherein each containerized environment houses one or more modules configured to perform specific transformation tasks on a visual digital asset, the specific transformation tasks comprising at least one of converting, optimizing, or rendering the visual digital asset;an input interface configured to receive a definition of an orchestration pipeline, the orchestration pipeline specifying a sequence of transformation tasks to be performed on the visual digital asset, each transformation task corresponding to a particular module within a particular containerized environment;a dynamic loader configured to dynamically load one or more pipeline modules based on the orchestration pipeline, wherein each pipeline module exposes an interface that allows the model operations orchestrator to provide specific configuration parameters required for execution of a corresponding transformation task for that pipeline module; a scheduler configured to schedule execution of the one or more pipeline modules in accordance with the orchestration pipeline, wherein the one or more pipeline modules are executed either sequentially or in parallel based on the sequence of transformation tasks and without collision; and a processing engine configured to process the visual digital asset through the sequence of transformation tasks as defined in the orchestration pipeline, wherein the visual digital asset is managed by the model operations orchestrator as a single manifest asset that is passed between the one or more pipeline modules and handled throughout the orchestration pipeline. (NOTE: Using Vasquez digital asset processing orchestration pipeline in processing the visual digital asset of Payne, and implementing Doshi’s barrier synchronization to prevent collision in the execution of the orchestration pipeline.)
The reason for doing so is for “the removal of external dependencies and automation where possible, and the re-use of modelled product components, products, scripts and environments to produce digital assets. Further improvement in automation speed can be achieved through use of purely aesthetic 3D modelling” (Payne: ¶8)
Regarding CRM claim 28, CRM claim 28 is drawn to the CRM corresponding to the functions of using same as claimed in the apparatus of claim 29. Therefore, CRM claim 28 corresponds to the functions in the apparatus of claim 29, and is rejected for the same reasons of obviousness as used above.
Regarding method claim 1, method claim 1 is drawn to the method corresponding to the functions of using same as claimed in apparatus claim 29. Therefore, method claim 1 corresponds to the operations in the apparatus of claim 29, and is rejected for the same reasons of obviousness as used above.
Regarding claim 2, depending on 1,
The combination of Vasquez, Doshi, and Payne teaches:
The method of claim 1,
wherein the sequence of transformation tasks are selected from a group consisting of: downloading the visual digital asset, uncompressing the visual digital asset, converting a format of the visual digital asset, reconverting a converted visual digital asset into an additional format, transforming the visual digital asset, optimizing the visual digital asset, sanitizing the visual digital asset, generating a thumbnail of the visual digital asset, compressing the visual digital asset, rendering the visual digital asset, gathering metrics on the visual digital asset, storing the visual digital asset, uploading the visual digital asset, and uploading a plurality of versions of the visual digital asset.
(Vasquez: col 16 lines 35-40, “. . . the DCAMPS-C container continually monitors the classification results and, when needed, updates the rules and policies. . .” col 17 lines 7-10, “. . . the data source integration is configured to allow the data classification engine to scan and classify data assets and content within and across a plurality . . . ” NOTE: Scanning and monitoring classification results is gathering metrics)
(Vasquez: col 16 lines 25-30, “. . . the step for generating enriched metadata for the digital assets and content is performed by the automated metadata publishing service as part of the metadata enrichment process. . .”; NOTE: the digital assets are transformed when the metadata is enriched).
(Vasquez: col 17 lines 15-23, “. . . the automated metadata publishing service is configured to receive the classified data and generated metadata from the data classification engine. After receiving the classified data and the generated metadata, the automated metadata publishing service of some embodiments is configured to perform a metadata enrichment process to add the metadata to the digital assets and content of the organization . . .” col 18 lines 63-65, “. . . the file server (at 220) is a system that is configured to store various files and documents in the organization. . .” ; NOTE: In order to perform metadata enrichment to add metadata to digital assets, it requires downloading the digital assets themselves first to be classified. The system then uploads and stores the updated or transformed digital asset to the file server. Download digital asset >> scan >> classify >> publish >> upload and store).
Regarding claim 3, depending on 1,
The combination of Vasquez, Doshi, and Payne teaches:
The method of claim 1,
wherein the visual digital asset is selected from a group consisting of: a static image, a two-dimensional image, a three-dimensional image, a holographic image, a video, a three-dimensional video, and a three-dimensional model (Vasquez: col 14 lines 1-7, “. . . the DCAMPS-C system can be adapted for use in different ways or in other fields of technology, including media asset management, whereby DCAMPS-C is adapted to classify and manage multimedia content, such as images, videos, and audio files . . .”)
Regarding claim 4, depending on 1,
The combination of Vasquez, Doshi, and Payne teaches:
The method of claim 1,
wherein a single visual digital asset input into the orchestration pipeline produces a plurality of configured outputs (Vasquez: col 3 lines 52-55, “. . . produce more accurate and comprehensive metadata. . .”; col 18 lines 52-55, “. . . The data classification engine (at 245) has a primary role of classifying the data and generating metadata. . .”; col 14 lines 9-11, “. . . When employed in managing product information (including product specifications, descriptions, and attributes). . .”; NOTE: A produced metadata from a digital asset is a configured output. The produced classification is also a configured output. The metadata and classification constitutes a plurality of configured outputs produced from a single digital asset. Additionally, when applied in managing product information, the product specification, description, and attributes produced from analyzing the digital asset also constitutes a plurality of configured outputs.)
Regarding claim 5, depending on 1,
The combination of Vasquez, Doshi, and Payne teaches:
The method of claim 1,
wherein the orchestration pipeline is a multi-use pipeline that can be reused for multiple visual digital assets by updating state values of variables within the orchestration pipeline before executing the orchestration pipeline (Vasquez: col 14 lines 1-5, “. . . the DCAMPS-C system can be adapted for use in different ways or in other fields of technology, including media asset management, whereby DCAMPS-C is adapted to classify and manage multimedia content, such as images, videos, and audio files. . .; col 16 lines 35-40, “. . . the DCAMPS-C container continually monitors the classification results and, when needed, updates the rules and policies. In this way, the DCAMPS-C container for the organization continuously improves model accuracy, and also improves accuracy via a feedback loop (at 175). . .”; NOTE: Vasquez system can be reused for multiple digital assets such as images, videos, and audio files. As discussed above, before executing the orchestration pipeline, the system translates the language input first to generate a formal rule for classification. Therefore, when the language input is received, the state of values of variables is updated to update and generate the formal rules. Additionally, DCAMPS-C container continually monitors the classification results, and when needed because the state of variables are updated, the system updates the rules and policies accordingly in a loop reusing the orchestration pipeline.)
Regarding claim 6, depending on 5,
The combination of Vasquez, Doshi, and Payne teaches:
The method of claim 5,
Vasquez further teaches: further comprising:
providing a plugin that iterates through all transformation tasks of the sequence of transformation tasks and adds user-defined adjustments to each transformation task (Vasquez: col 6 lines 39-44, “. . . the DCAMPS-C system further comprises runtime program components comprising program code (software), runtime executables, plug-in components, system tools, libraries, and settings that provide integral components for runtime operation as a standalone, executable package for a single instance of the DCAMPS-C container. . .”; NOTE: As discussed in the rejection of claim 5, DCAMPS system updates the rules and policies via a feedback loop to improve classification accuracy, when it loops, it reiterates through the sequence of tasks. The DCAMPS system also comprise of plug-in component that is integrated to the DCAMPS system, therefore, it also iterates through the sequence of tasks. A user-defined adjustment is a language input from a user. When the input changes, the generated formal rules update, then the system will adjust accordingly based on the updated formal rules.).
Regarding claim 7, depending on 1,
The combination of Vasquez, Doshi, and Payne teaches:
The method of claim 1,
Vasquez further teaches:
wherein the orchestration pipeline is generated based on a user input (Vasquez: col 7 lines 59-62, :. . . YOLAGPT leverages state-of-the-art NLP techniques and advanced language models, thereby enabling natural language input (by end users) for defining and managing classification rules. . .”NOTE: As discussed in the rejection of claim 29, the orchestration pipeline is configured based on formal rules for classification. The formal rules are based on interpreting the language input from a user.).
Regarding claim 8, depending on 7,
The combination of Vasquez, Doshi, and Payne teaches:
The method of claim 7,
Although Vasquez teaches input devices for users to interact with including voice assistant devices and touchscreen devices with graphical interface elements to generate and configuring the orchestration pipeline, and monitoring the orchestration pipeline, Vasquez does not disclose providing an interactive drag-and-drop user interface to receive the user input for designing, monitoring, and managing the orchestration pipeline, wherein various functions, features, or operations are graphically represented as blocks from available libraries, which can be dragged into a user interface and interconnected in a particular order to establish the orchestration pipeline, with variables entered through customized configurations.
The analogous art Doshi teaches:
further comprising:
providing an interactive drag-and-drop user interface to receive the user input for designing, monitoring, and managing the orchestration pipeline, wherein various functions, features, or operations are graphically represented as blocks from available libraries, which can be dragged into a user interface and interconnected in a particular order to establish the orchestration pipeline, with variables entered through customized configurations (Doshi: ¶219, “Information design principles may use “property graphs” . . . consisting of nodes and edges may have properties associated with either nodes or edges. . . The authoring interface 4200 applies property graph design principles to represent security constructs in a “drag-and-drop” property graph abstraction. . . translates user input into workload functions, data, and orchestration operations . . . workload module A or workload module B to present authentication, authorization, attestation, or other credentials as a pre-requisite to WL information exchange. . . ” ¶64, “FIG. 3 is a block diagram illustrating the components of a PaaS layer (e.g., corresponding to layer 104), according to an example. Here, a PaaS layer 302 includes a variety of APIs, functions . . . operably coupled with a user (e.g., services user) dashboard UI 304 . . . to invoke use of these APIs, functions, and services. The features discussed herein introduce the use of trusted C2E capabilities 310 for the PaaS layer, which include but are not limited to: node feature discovery; state change monitoring; attestation; telemetry; trust policy management; and trust services . . .” ;NOTE: The blocks correspond to the functions in a PaaS Layer library shown in Figure. 3 that users can drag and drop in particular order to translate user input into workload functions establishing orchestration pipeline operations.)
It would have been obvious to a person having ordinary skill in the art (PHOSITA) before the effective filing date of the claimed invention to combine Vasquez, Doshi, and Payne and include: providing an interactive drag-and-drop user interface to receive the user input for designing, monitoring, and managing the orchestration pipeline, wherein various functions, features, or operations are graphically represented as blocks from available libraries, which can be dragged into a user interface and interconnected in a particular order to establish the orchestration pipeline, with variables entered through customized configurations.
The reason for doing so is to provide users with easy to use drag and drop coding for designing an orchestration pipeline as taught by Doshi without requiring manual coding as the blocks and functions are readily available for the user. In addition, to apply “property graph design principles to represent security constructs in a “drag-and-drop” property graph abstraction” (Doshi: ¶219)
Regarding claim 9, depending on 7,
The combination of Vasquez, Doshi, and Payne teaches:
The method of claim 7,
Although Vasquez teaches generating orchestration pipeline based on user input and feedback loop, Vasquez is not explicit if it stores the generated orchestration pipeline configurations as templates in a repository.
Doshi further teaches:
providing a repository of pre-defined templates for a set of orchestration pipelines for processing visual digital assets (Doshi: ¶91, “FIG. 11 is a block diagram illustrating a classic workload testing infrastructure, . . . test framework 1106 is then used to generate respective scripts for a system under test 1108. The test scenario is repeatable, as WL execution follows a predetermined flow that uses resources . . . ”; NOTE: The templates in Fig. 11 ref 1106 is the repository of pre-defined templates, which are reusable scripts to setup an orchestration pipelines for processing digital assets. Additionally, in reference to the rejection of claim 8, the PaaS layer as shown in Fig. 3 is a repository of pre-defined templates for functions to generate an orchestration pipeline.)
It would have been an obvious design choice between saving/not saving the orchestration pipeline configuration in a repository to a person having ordinary skill in the art (PHOSITA) before the effective filing date of the claimed invention and combine Vasquez, Doshi, and Payne to store generated orchestration pipeline configurations providing a repository of pre-defined templates for a set of orchestration pipelines for processing digital assets.
The reason for doing so is provide reusability of previously generated orchestration pipeline configuration to save time for task requiring the same configuration without creating the configuration from scratch predictably saving the user time and resources.
Regarding claim 10, depending on 1,
The combination of Vasquez, Doshi, and Payne teaches:
The method of claim 1,
Vasquez further teaches:
further comprising:
allowing nodes to be introduced or removed from the orchestration pipeline without disruption to existing executions (Vasquez: col 2 lines 40-45, “. . . a DCAMPS orchestrator that is configured to manage deployment of the DCAMPS-C container and scaling of the components encapsulated in the deployed DCAMPS-C container. . .”; col 12 lines 45-50, “. . . the DevOps and cloud-native infrastructure integration logic may be based on an implementation of best practices or may involve use of tools and/or workflows that enable seamless deployment, scaling, and management of DCAMPS-C in modern IT environments. . .”; NOTE: Scaling constitute to introducing nodes (encapsulated components) when scaled higher, or removing nodes when scaled lower. Since the system enables seamless deployment and scaling, therefore the introduction and removal of nodes does not disrupt the existing orchestration pipeline execution).
Regarding claim 11, depending on 1,
The combination of Vasquez, Doshi, and Payne teaches:
The method of claim 1,
Vasquez further teaches:
further comprising: receiving a request to integrate a custom node into the orchestration pipeline (NOTE: in reference to the rejection claim 29, the user input is the request. Once the request is received, the custom nodes such as the LLM container and DCAMPS-C container are executed and integrated into the orchestration pipeline);
However, Vasquez fails to teach: performing a compatibility pre-analysis on the custom node before integrating the custom node into the orchestration pipeline.
Doshi further teaches:
performing a compatibility pre-analysis on the custom node before integrating the custom node into the orchestration pipeline (Doshi: ¶216, “. . . if a workload node is dragged-and-dropped from outside the security boundary to inside the security boundary, an onboarding sub-workflow is automatically applied such that the ingressing node becomes subject to attestation, authentication, audit, and/or authorization challenges . . .”¶219, “Information design principles may use “property graphs” as a term of art meaning a graph consisting of nodes and edges may have properties associated with either nodes or edges. The authoring interface 4200 applies property graph design principles to represent security constructs in a “drag-and-drop” property graph abstraction. The authoring interface 4200 translates user input into workload functions, data, and orchestration operations such that the A-to-B interaction is instantiated in a workload by applying a secure challenge/response protocol (e.g., TLS, HTTPS, SPDM, IPSEC, CoAPS) that requires either workload module A or workload module B to present authentication, authorization, attestation, or other credentials as a pre-requisite to WL information exchange. This may be automatically applied to the WL definition or may be applied using a previously designed secure challenge/response workflow subroutine. . .”; NOTE: The compatibility pre-analysis is the node being subjected to attestation, authentication, authorization challenges before the node is integrated into the orchestration pipeline as a prerequisite to workload information exchange. There is an incompatibility when the attestation, or authentication is not satisfied).
Regarding claim 12, depending on 1,
The combination of Vasquez, Doshi, and Payne teaches:
The method of claim 1,
Vasquez further teaches:
wherein the orchestration pipeline is generated based on an artificial intelligence input programmed to generate optimized pipelines based on specific attributes and requirements of the visual digital asset (Vasquez: col 2 lines 5-30, “. . . an automated metadata publishing service that is configured to receive the classified data and generated metadata from the data classification engine and perform a metadata enrichment process to add the metadata to the digital assets . . . a “your optimized learning artificial-intelligence generative pre-trained transformer” (YOLAGPT) large language model (LLM) module that is configured to provide . . . classification of data assets and content, . . . the YOLAGPT LLM module provides seamless interpretation of natural language inputs and classification of data assets and content by . . .scanning and classifying data assets and content across the data sources of the organization based on the classification rules. . .”).
Regarding claim 13, depending on1,
The combination of Vasquez, Doshi, and Payne teaches:
The method of claim 1,
Vasquez further teaches:
further comprising:
configuring the orchestration pipeline based on a type of the visual digital asset (Vasquez: col 14 lines 1-5, “. . . the DCAMPS-C system can be adapted for use in different ways or in other fields of technology, including media asset management, whereby DCAMPS-C is adapted to classify and manage multimedia content, such as images, videos, and audio files. . .”; NOTE: When the system adapts to a type of digital asset (images, videos, audio), it is configured to classify and process the digital asset with a specific type.).
Regarding claim 14, depending on 1,
The combination of Vasquez, Doshi, and Payne teaches:
The method of claim 1,
Although Vasquez teaches: dynamically allocating resources when scaling introducing nodes (NOTE: When adding nodes or containers for scaling, resources are allocated inherently, also see the rejection of claim 10), Vasquez fails to explicitly disclose of the resource allocation is based on a complexity and demand of the sequence of tasks and a nature of the digital asset.
Doshi teaches:
dynamically allocating resources to nodes based on a complexity and demand of the sequence of transformation tasks and a nature of the visual digital asset (Doshi: ¶157, “Infrastructure nodes may be hosted by an infrastructure processor mesh . . . IPU pools may be connected for efficient resource sharing, load balancing, and resiliency, including for respective resource slices 3220”; ¶166, “WLs may be dynamically partitioned, distributed and deployed to be highly responsive and to draw on federated infrastructure for execution, storage, acceleration, and power”; ¶120, “. . . workload can be partitioned and distributed. Thus, the precise resource requirements for a WL partition are readily available to orchestrators or cluster resource managers for dynamic distribution or coalescence operations. . .”; ¶71, “Workload authoring may also include elastic WL properties that aid in decomposing a monolithic WL into distributable parts using metadata . . .,” NOTE: Workloads = WLs. Load balancing is allocating more resource to the node which demands more processing. The value of the precise resource requirement is the amount of complexity of the sequence of task, if the sequence of task is not high, then the resource requirement is low, and when the complexity of the sequence of task, then the resource requirement is high. A metadata includes information that defines a nature of a digital asset. Therefore, Doshi’s resource allocation is based on complexity and demand and nature of a digital asset.)
It would have been obvious to a person having ordinary skill in the art (PHOSITA) before the effective filing date of the claimed invention to combine Vasquez, Doshi, and Payne and include: dynamically allocating resources to nodes based on a complexity and demand of the sequence of tasks and a nature of the digital asset.
The reason for doing so is to provide a “precise resource requirements for a WL partition are readily available to orchestrators or cluster resource managers for dynamic distribution or coalescence operations” (Doshi: ¶120)
Regarding claim 15, depending on 1,
The combination of Vasquez, Doshi, and Payne teaches:
The method of claim 1,
Although Vasquez teaches scaling of the orchestration pipeline, Vasquez fails to teach forecasting resource requirements for the model operations orchestrator based on patterns, asset types, and technology evolution to enable proactive system scaling and infrastructure planning.
Doshi further teaches
further comprising:
forecasting resource requirements for the model operations orchestrator based on patterns, asset types, and technology evolution to enable proactive system scaling and infrastructure planning (Doshi: ¶84, “The elastic WL configuration enables performance improvements while being cyber-resilient. Elastic WL telemetry and AI/ML algorithms may add greater context for leveraging WL archives that quantify WL lifecycle overhead . . . may record resource utilization, latency, network bandwidth and so forth. Thus, the overhead that is associated with WL lifecycle transitions can be analyzed, when these measurements are available for analysis and optimization of WL microservices”; ¶89-91, “. . . risks are particularly possible as new or updated elastic workloads are introduced . . . workload simulation building blocks with a simulated infrastructure resource supply and elastic intents. These intents can be used to enable multiple systems under test to compete for a simulated finite set of resources as elastic intents are used to direct workloads to fragment and distribute across the simulated resource infrastructure. Multiple distribution and coalescence scenarios can be exercised while also manipulating resource availability profiles that fully exploit the effects of elasticity for a given set of elastic workloads. This also enables testing of elastic workload behaviors in simulation before subjecting them to production deployments where risks of misbehavior are multiplied”; NOTE: Doshi’s system analyzes lifecycle transitions, which are patterns; simulates workload anticipating resource requirements because of risk of new or updated workloads, which is technology evolution; and in reference to the rejection of claim 14, Doshi’s system allocates resources based on nature of the digital asset, which is the type of asset. Anticipating resource requirements is forecasting. Doshi’s system is capable of forecasting resource requirements for the intended use to enable proactive system scaling and infrastructure planning).
It would have been obvious to a person having ordinary skill in the art (PHOSITA) before the effective filing date of the claimed invention to combine Vasquez, Doshi, and Payne and include forecasting resource requirements for the model operations orchestrator based on patterns, asset types, and technology evolution to enable proactive system scaling and infrastructure planning.
The reason for doing so is for “testing of elastic workload behaviors in simulation before subjecting them to production deployments where risks of misbehavior are multiplied” (Doshi: ¶90).
Regarding claim 16, depending on 1,
The combination of Vasquez, Doshi, and Payne teaches:
The method of claim 1,
Although Vasquez teaches a system that ensures accurate identification and categorization of sensitive information (NOTE: An asset that is identified and categorized to have sensitive information is a critical digital asset), Vasquez fails to teach providing prioritized execution control through task prioritization, ensuring that critical digital assets are processed first.
The combination further teaches:
providing prioritized execution control through task prioritization, ensuring that critical visual digital assets are processed first (Doshi: ¶100, “. . . Security vulnerability events (e.g., Common Vulnerabilities and Exposures (CVEs)) . . .”; ¶121, “. . . the elastic workload infrastructure may be diverse in terms of hardware and software BOM that are used. The CVEs may be prioritized based on rank (e.g., ranking of an impact if compromised for a given vulnerability) and this rank may assist in the upgrade strategy that is applied across the orchestration and deployment infrastructure. This may leverage residual compute resources to minimize operational runtime impact . . . A small amount of partitioned capacity may be reserved in respective nodes so that the workload can be migrated to its optimized partitioned capacity first (without having to drain the workload or migrate it long distance), and then restored to full capacity. . . after a CVE fix is applied and rebooted. . .”; NOTE: During orchestration and processing of digital assets, security vulnerabilities (CVEs) are identified which is critical digital assets. The CVE impact is ranked and prioritized for execution control through task prioritization when applied across the orchestration and deployment infrastructure. Prioritizing to fix the digital asset CVE ensures that critical digital assets are processed first).
It would have been obvious to a person having ordinary skill in the art (PHOSITA) before the effective filing date of the claimed invention to combine Vasquez, Doshi, and Payne and include providing prioritized execution control through task prioritization, ensuring that critical visual digital assets are processed first.
The reason for doing so is to leverage residual compute resources to minimize operational runtime impact (Doshi: ¶121) during the orchestration execution..
Regarding claim 17, depending on 1,
The combination of Vasquez, Doshi, and Payne teaches:
The method of claim 1,
However Vasquez fails to teach: determining whether a first task of the sequence of transformation tasks needs to wait for a second task of the sequence of transformation tasks to finish or if the first task and the second task can be run concurrently; and orchestrating the first task and the second task based thereon.
The combination further teaches:
determining whether a first task of the sequence of tasks needs to wait for a second task of the sequence of tasks to finish or if the first task and the second task can be run concurrently; and orchestrating the first task and the second task based thereon (Doshi: ¶61, “. . . The workflow may divide WL computations into execution operations that are in a particular sequence or in a particular concurrence . . .”; ¶168, “A barrier mechanism may be used to synchronize workloads. For instance, a barrier may pause, suspend, sleep, or put a workload in a wait state to allow it to synchronize with one or more other workloads that are going to rendezvous at the barrier.”; ¶207, “. . . rendezvous points in a task's sequence of operations, instructions, etc. occur in the form of synchronization primitives like Event-Wait( . . . parameters . . . ), Semaphore-Wait ( . . . parameters . . . ), Condition-Wait ( . . . parameters . . . ), and so on. . .”); ¶81, “. . . A hierarchical lock structure may be used to guarantee sequential and parallel executions can occur simultaneously while still maintaining overall integrity of the elastic workload. . .”; ¶177, “The barrier synchronization manager (BSM) is placed on universally available and reachable nodes so that the BSM can be a reliable anchor for receiving and acting on barrier check ins. The BMS may execute a boundary action upon receiving the last waited-for check in . . .”; ¶224, “. . . the method 4300 includes initiating execution of a second set of tasks executing . . .in response to determining that the barrier synchronization condition is satisfied . . .”; NOTE: As also referenced in the rejection of claim 29, Doshi’s barrier prevents collision and race conditions. Doshi’s hierarchical lock structure allows for simultaneous sequential and parallel executions. The determination whether to execute in sequential or parallel is based on the BSM receiving the last waited-for check in. The barrier then puts a first task to wait for a second task to finish to synchronize, otherwise, if there is no barrier check in, then run on parallel execution until a barrier check in is received.
The reason for doing so is prevent race condition and data corruption because “pipelined execution can exist where the input of one node in an elastic WL cluster is satisfied by the output of another cluster node forming a chain of operations that execute in sequence” (Doshi: ¶81).
Regarding claim 18, depending on 1,
The combination of Vasquez, Doshi, and Payne teaches:
The method of claim 1,
Vasquez teaches:
adjusting execution of the orchestration pipeline according to the status updates (Vasquez: col 4 lines 28-30, “. . . monitoring, by the DCAMPS-C container, classification results and updating rules and policies as needed, and (xv) continuously improving model accuracy through a feedback loop. . .”; NOTE: Status updates are identified from monitoring the classification results, it is then fed back to update the rules and policies resulting in an adjust execution of the orchestration pipeline).
However, Vasquez fails to teach receiving, by the model operations orchestrator, status updates from the one or more pipeline modules, including completion notifications and error messages
The combination further teaches:
receiving, by the model operations orchestrator, status updates from the one or more pipeline modules, including completion notifications and error messages (Doshi: ¶86, “. . . After successful application of the update (or error state), the PUM 1022 notifies the Workload Update Manager (WUM 1012) regarding the update status (Operation 4). The WUM 1012 may further notify an Orchestrator where notifications from across the elastic edge may be assessed. . .”; NOTE: notification of error state is an error message, notification for successful update is completion notification);
It would have been obvious to a person having ordinary skill in the art (PHOSITA) before the effective filing date of the claimed invention to combine Vasquez, Doshi, and Payne and include: receiving, by the model operations orchestrator, status updates from the one or more pipeline modules, including completion notifications and error messages.
The reason for doing so is to predictably notify the user for any errors or status updates during the execution of the orchestration pipeline so that errors can be fixed in a timely manner. Additionally, is so that “the Orchestrator nevertheless maintains a central and consistent view of update application using notifications and the WL metadata” (Doshi: ¶86).
Regarding claim 19, depending on 1,
The combination of Vasquez, Doshi, and Payne teaches:
The method of claim 1,
However Vasquez fails to teach providing global callbacks to the orchestration pipeline for tracking task metrics.
The combination further teaches:
providing global callbacks to the orchestration pipeline for tracking transformation task metrics (Doshi: ¶84, “The elastic WL configuration enables performance improvements while being cyber-resilient. Elastic WL telemetry and AI/ML algorithms may add greater context for leveraging WL archives that quantify WL lifecycle overhead (e.g., pause/restart vs. decommission/recommission). For example, the clastic WL equivalent of a suspend-resume state “S3” in an operating system process suspend/resume may record resource utilization, latency, network bandwidth and so forth. Thus, the overhead that is associated with WL lifecycle transitions can be analyzed, when these measurements are available for analysis and optimization of WL microservices”; NOTE: telemetry is the global call back to the orchestration pipeline for recording WL task metrics such as resource utilization, latency, network bandwidth.).
It would have been obvious to a person having ordinary skill in the art (PHOSITA) before the effective filing date of the claimed invention to combine Vasquez, Doshi, and Payne and include: providing global callbacks to the orchestration pipeline for tracking transformation task metrics.
The reason for doing so is because “the overhead that is associated with WL lifecycle transitions can be analyzed, when these measurements are available for analysis and optimization of WL microservices” (Doshi: ¶84).
Regarding claim 20, depending on 1,
The combination of Vasquez, Doshi, and Payne teaches:
The method of claim 1,
However, Vasquez fails to teach providing real-time analytics of processing status, resource utilization, and performance metrics correlated to particular types of digital assets.
The combination further teaches:
providing real-time analytics of processing status, resource utilization, and performance metrics correlated to particular types of visual digital assets (Doshi: ¶84, “The elastic WL configuration enables performance improvements while being cyber-resilient. Elastic WL telemetry and AI/ML algorithms may add greater context for leveraging WL archives that quantify WL lifecycle overhead (e.g., pause/restart vs. decommission/recommission). For example, the clastic WL equivalent of a suspend-resume state “S3” in an operating system process suspend/resume may record resource utilization, latency, network bandwidth and so forth. Thus, the overhead that is associated with WL lifecycle transitions can be analyzed, when these measurements are available for analysis and optimization of WL microservices”; ¶86, “. . . After successful application of the update (or error state), the PUM 1022 notifies the Workload Update Manager (WUM 1012) regarding the update status (Operation 4). The WUM 1012 may further notify an Orchestrator where notifications from across the elastic edge may be assessed”; ¶109, “The top-half update package 1504 also may contain updated workload code, settings, configuration parameters, quality-of-service policies, key performance indicators, metrics, data source, data sink parameters, etc. It may contain attestation information in the form of reference values, endorsed values, device identity certificates and other information used to attest the workload and to verify peer workloads”; NOTE: A notification a successful application of the update or error state is a real-time analytics of processing status. Resource utilization is provided for analysis. Because the telemetry is tracking metrics from the workload that’s executing a particular digital asset, then the performance indicators or metrics are correlated to that particular type of digital asset being processed in addition to the information in the bill of materials as disclosed in paragraph 108 which is directly related to the digital asset being processed.).
It would have been obvious to a person having ordinary skill in the art (PHOSITA) before the effective filing date of the claimed invention to combine Vasquez, Doshi, and Payne and include: providing real-time analytics of processing status, resource utilization, and performance metrics correlated to particular types of digital assets.
The reason for doing so is to predictably provide users with status information of the orchestration pipeline execution such as real-time notification of errors and performance indicator to confirm that the orchestration pipeline execution functions as intended. Additionally, is so that “the overhead that is associated with WL lifecycle transitions can be analyzed, when these measurements are available for analysis and optimization of WL microservices” (Doshi: ¶84).
Regarding claim 21, depending on 1,
The combination of Vasquez, Doshi, and Payne teaches:
The method of claim 1,
However, Vasquez fails to teach: providing a centralized logging mechanism that consolidates logs from all containers within the orchestration pipeline.
Doshi further teaches:
providing a centralized logging mechanism that consolidates logs from all containers within the orchestration pipeline (Doshi: ¶109, “The top-half update package 1504 also may contain updated workload code, settings, configuration parameters, quality-of-service policies, key performance indicators, metrics, data source, data sink parameters, etc. It may contain attestation information in the form of reference values, endorsed values, device identity certificates and other information used to attest the workload and to verify peer workloads”; ¶293, “. . . Additional features of FaaS may include: a . . . common data storage to store data for reuse by one or more functions; orchestration and management among individual functions; function execution management, parallelism, and consolidation; management of container and function memory spaces; coordination of acceleration resources available for functions; and distribution of functions between containers (including “warm” containers, already deployed or operating, versus “cold” which require deployment or configuration).”; NOTE: The top-half package logs various information tracked during orchestration pipeline execution. FaaS is the centralized logging mechanism which include a common data storage to store and consolidate data from managed containers. Additionally, there is a function block labeled “logging” available to users as shown in Fig. 3.).
It would have been obvious to a person having ordinary skill in the art (PHOSITA) before the effective filing date of the claimed invention to combine Vasquez, Doshi, and Payne and include: providing a centralized logging mechanism that consolidates logs from all containers within the orchestration pipeline.
The reason for doing so is to provide a predictable common storage to store data or logs for easy access and makes searching more efficient. Additionally “so that updates to a persistent state can be kept traceable and auditable” (Doshi: ¶191).
Regarding claim 22, depending on 1,
The combination of Vasquez, Doshi, and Payne teaches:
The method of claim 1,
However, Vasquez fails to teach: detecting an encountered problem within execution of the orchestration pipeline; determining a responsive action for the encountered problem selected from a group consisting of: retrying a given task, skipping the given task, diverting the given task to another similar node if available, and reporting the encountered problem; and performing the responsive action.
The combination further teaches:
detecting an encountered problem within execution of the orchestration pipeline (Doshi: ¶86, “. . . After successful application of the update (or error state), the PUM 1022 notifies the Workload Update Manager (WUM 1012) regarding the update status (Operation 4). The WUM 1012 may further notify an Orchestrator 1010 where notifications from across the elastic edge may be assessed (Operation 5). Distributing WL update operations improves scalability of elastic (distributed) WLs because the distributed WL does not need to be coalesced to a central node to apply an update then re-distribute the WL. The Orchestrator nevertheless maintains a central and consistent view of update application using notifications and the WL metadata. . .”; NOTE: The error state is an encountered problem within the execution of the orchestration pipeline. Once error state is detected, a notification is generated for assessment.);
determining a responsive action for the encountered problem selected from a group consisting of: retrying a given transformation task, skipping the given transformation task, diverting the given transformation task to another similar node if available, and reporting the encountered problem; (NOTE: The determined responsive action is the system generates an error state notification, which constitute to reporting the encountered problem.)
and performing the responsive action (NOTE: When the notification is generated, then the responsive action of reporting the encountered problem is performed.).
(Doshi: ¶86, “. . . After successful application of the update (or error state), the PUM 1022 notifies the Workload Update Manager (WUM 1012) regarding the update status (Operation 4). The WUM 1012 may further notify an Orchestrator 1010 where notifications from across the elastic edge may be assessed (Operation 5). Distributing WL update operations improves scalability of elastic (distributed) WLs because the distributed WL does not need to be coalesced to a central node to apply an update then re-distribute the WL. The Orchestrator nevertheless maintains a central and consistent view of update application using notifications and the WL metadata. . .”).
It would have been obvious to a person having ordinary skill in the art (PHOSITA) before the effective filing date of the claimed invention to combine Vasquez, Doshi, and Payne and include: detecting an encountered problem within execution of the orchestration pipeline; determining a responsive action for the encountered problem selected from a group consisting of: retrying a given transformation task, skipping the given transformation task, diverting the given transformation task to another similar node if available, and reporting the encountered problem; and performing the responsive action.
The reason for doing so is to provide a predictable notification when an error is identified such that “the Orchestrator nevertheless maintains a central and consistent view of update application using notifications and the WL metadata” (Doshi: ¶86)
Regarding claim 23, depending on 1,
The combination of Vasquez, Doshi, and Payne teaches:
The method of claim 1,
However, Vasquez fails to teach: analyzing historical data and patterns of the orchestration pipeline over time; and using predictive analysis based on analyzing to adjust the orchestration pipeline.
Doshi teaches:
analyzing historical data and patterns of the orchestration pipeline over time; and using predictive analysis based on analyzing to adjust the orchestration pipeline (Doshi: ¶84, “The elastic WL configuration enables performance improvements while being cyber-resilient. Elastic WL telemetry and AI/ML algorithms may add greater context for leveraging WL archives that quantify WL lifecycle overhead (e.g., pause/restart vs. decommission/recommission). For example, the clastic WL equivalent of a suspend-resume state “S3” in an operating system process suspend/resume may record resource utilization, latency, network bandwidth and so forth. Thus, the overhead that is associated with WL lifecycle transitions can be analyzed, when these measurements are available for analysis and optimization of WL microservices.”; NOTE: The WL archives is the historical data and the lifecycle transitions is the patterns. They are analyzed to optimize the WL microservices adjusting the orchestration pipeline. The analysis is predictive because it uses AI/ML to predict optimization parameter based on the analyzed WL archives and lifecycle overheads.)
It would have been obvious to a person having ordinary skill in the art (PHOSITA) before the effective filing date of the claimed invention to combine Vasquez, Doshi, and Payne and include: analyzing historical data and patterns of the orchestration pipeline over time; and using predictive analysis based on analyzing to adjust the orchestration pipeline.
The reason for doing so that “the overhead that is associated with WL lifecycle transitions can be analyzed, when these measurements are available for analysis and optimization of WL microservices” (Doshi: ¶84).
Regarding claim 24, depending on 23,
The combination of Vasquez, Doshi, and Payne teaches:
The method of claim 23,
However, Vasquez fails to teach wherein adjusting the orchestration pipeline comprises one or more of: preemptively scheduling tasks, optimizing resource allocation, forecasting potential bottlenecks, and forecasting system stresses.
The combination teaches:
wherein adjusting the orchestration pipeline comprises one or more of:
preemptively scheduling transformation tasks, optimizing resource allocation, forecasting potential bottlenecks, and forecasting system stresses
(Doshi: ¶118, “. . . A proactive migration 1704 may apply to top-half nodes as well. A new migration target 1710A (guest or secure enclave) is created with a clone of the guest/enclave to be migrated, and then the update 1706 is applied. The migration target 1710A creates a new connection to the security-services module 1712 which multiplexes with the IO devices. The old enclave/guest migrates workload and data to the migration target . . .”; NOTE: The migration task is pre-emptively scheduled before the update).
(Doshi: ¶99, “. . . the resource simulation framework . . . The simulation objective then can be performed to optimize resource allocation . . .”; NOTE: optimizing resource allocations)
It would have been obvious to a person having ordinary skill in the art (PHOSITA) before the effective filing date of the claimed invention to combine Vasquez, Doshi, and Payne and include: wherein adjusting the orchestration pipeline comprises one or more of: preemptively scheduling transformation tasks, optimizing resource allocation, forecasting potential bottlenecks, and forecasting system stresses.
The reason for doing so to “provide a cloud-like distributed service that offers orchestration and management for applications among many types of storage and compute resources” (Doshi: ¶2).
Regarding claim 25, depending on 1,
The combination of Vasquez, Doshi, and Payne teaches:
The method of claim 1,
Vasquez further teaches:
refining the orchestration pipeline according to a self-optimization mechanism based on assessment of previous results from the orchestration pipeline (Vasquez: col 4 lines 28-30, “monitoring, by the DCAMPS-C container, classification results and updating rules and policies as needed, and (xv) continuously improving model accuracy through a feedback loop”; NOTE: The orchestration pipeline is refined once it is improved. The self-optimization mechanism is the feedback loop. The assessment of previous classification results updates the rules that is fed back for continuous improvement or refinement.).
Regarding claim 26, depending on 25,
The combination of Vasquez, Doshi, and Payne teaches:
The method of claim 25,
Vasquez further teaches:
conducting an artificial intelligence-driven root-cause analysis to prevent future occurrences of an experienced error (Vasquez: col 2 lines 10-17, “. . . a “your optimized learning artificial-intelligence generative pre-trained transformer” (YOLAGPT) large language model (LLM) module that is configured to provide seamless interpretation of natural language inputs and classification of data assets and content . . .”; col 4 lines 28-30, “monitoring, by the DCAMPS-C container, classification results and updating rules and policies as needed, and (xv) continuously improving model accuracy through a feedback. . .”; (NOTE: The YOLAGPT is an AI GPT that conducts root-cause analysis by monitoring the classification results. Vasquez feedback mechanism improves accuracy continuously based on the classification results. In order to improve accuracy, an error or classification error is identified so the LLM can improve accuracy, otherwise if error is not detected, then updating the rules and policies is not needed. Once the accuracy of the classification is improved, then it prevents future occurrences of that classification error.)
Regarding claim 27, depending on 1,
The combination of Vasquez, Doshi, and Payne teaches:
The method of claim 1,
wherein the sequence of transformation tasks are executed on local environments, cloud environments, or a combination of both (Vasquez: Abstract, “A containerized data classification and automated metadata publishing system (DCAMPS-C) is disclosed. DCAMPS-C is powered by a data classification engine. The containerized data classification and automated metadata publishing system hosts a data classification and automated metadata cloud application publishing service. DCAMPS-C offers a containerized solution that automates the data classification and metadata generation processes, eliminates the need for manual input, reduces the chances of human error, and ensures seamless integration with modern cloud and on-premise infrastructures for better resource utilization, scalability, and overall efficiency in managing digital content within an organization”).
Conclusion
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any 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 PATRICK GALERA whose telephone number is (571)272-5070. The examiner can normally be reached Mon-Fri 0800-1700 ET.
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, King Poon can be reached at 571-270-0728. 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.
/PATRICK P GALERA/Examiner, Art Unit 2617 /KING Y POON/Supervisory Patent Examiner, Art Unit 2617