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 .
This action is responsive to the Amendment filed on February 25, 2026. Claims 1, 8, 10, 12, 14, and 15 are amended. Claims 1-20 are pending in the case. Claims 1, 8, and 15 are the independent claims.
This action is final.
Applicant’s Response
In the Amendment filed on February 25, 2026, Applicant amended the claims and provided arguments in response to the rejections of the claims under 35 USC 101 and 103 in the previous office action.
Response to Argument/Amendment
Applicant’s amendments to the claims in response to the rejection of the claims under 35 USC 101 in the previous office action are acknowledged, and Applicant’s associated arguments have been fully considered. Applicant’s amendments to the claims render the rejection moot. Therefore, the rejection is withdrawn.
Applicant’s amendments to the claims in response to the rejection of the claims under 35 USC 103 in the previous office action are acknowledged, and Applicant’s associated arguments have been fully considered. Applicant argues that Kim does not disclose “starting, by an integration layer on a cloud platform, a webserver to accept requests” and “scanning, by the integration layer, an application environment for a model loading function for an artificial intelligence/machine learning (AI/ML) model and a model invocation function for the AI/ML model.”
Regarding the limitation “starting, by an integration layer on a cloud platform, a webserver to accept requests,” as cited in the previous office action, Kim teaches starting, on a cloud platform, a webserver to accept requests (e.g. paragraph 0075, client 400 requesting device 100 for artificial intelligence service, and device 100 providing artificial intelligence service to the client 400; paragraph 0076, Fig. 2, electronic device 100 implemented as a server such as a cloud server). That is, as discussed in the cited portions of Kim, the device 100 may be implemented as a server which receives requests from client devices for artificial intelligence services, and which further responds to those requests by providing the requested services. Therefore, Kim appears to teach this limitation, i.e. because the device 100, implemented as a server, clearly is able to receive and respond to requests by providing a requested service, the server has clearly been started, for the purpose of accepting requests (i.e. where providing a requested service indicates acceptance of the request for the service). Moreover, as cited in the previous office action, Coven teaches that this step may be performed by an integration layer (e.g. col. 7 lines 51-54, data processing technologies offered as data processing services via API; col. 8 lines 15-61, integration layer triggering execution of data processing services and receiving outputs of data processing services, which are plugged-in to the integration layer; invocation of data processing service triggered by plug-ins of the integration layer; plug-in facilitates communication between cloud based services and hosts of data processing services; plug-ins registered with the integration layer; registration of data processing service with integration layer which in turn may register existence of data processing service with cloud based collaboration platform; registration process includes registration of domain of host of data processing service, registration of mechanism to invoke the data processing service from the integration layer, registration of function of data processing service, registration of characteristics of inputs to the data processing service, registration of outputs of the data processing service, etc.; col. 11 lines 1-5, “skill” refers to a module that handles processing of content using data processing technologies; col. 20 lines 32-36, indicating that skills module may include/be implemented using machine learning models; i.e. a skill, which is a data processing technology/service, may be implemented using an AI/ML model, and may be registered for use/invocation via an integration layer which then triggers execution of the corresponding services (implemented via the model), etc. in response to user requests, etc.). That is, Coven clearly teaches that the service (i.e. provided by a server, as taught by Kim) may be triggered/invoked (i.e. started as recited in the claims) by an integration layer. Therefore, this argument is not persuasive.
Regarding the limitation “scanning, by the integration layer, an application environment for a model loading function for an artificial intelligence/machine learning (AI/ML) model and a model invocation function for the AI/ML model,” Applicant’s argument is persuasive, and the rejection is withdrawn.
New grounds of rejection are provided below.
Claim Rejections – 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
The factual inquiries set forth in Graham v. John Deere Co., 383 U.S. 1, 148 USPQ 459 (1966), that are applied for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
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.
This application currently names joint inventors. In considering patentability of the claims under pre-AIA 35 U.S.C. 103(a), the examiner presumes that the subject matter of the various claims was commonly owned at the time any inventions covered therein were made absent any evidence to the contrary. Applicant is advised of the obligation under 37 CFR 1.56 to point out the inventor and invention dates of each claim that was not commonly owned at the time a later invention was made in order for the examiner to consider the applicability of pre-AIA 35 U.S.C. 103(c) and potential pre-AIA 35 U.S.C. 102€, (f) or (g) prior art under pre-AIA 35 U.S.C. 103(a).
Claims 1, 5-8, 12-15, 19, and 20 are rejected under 35 U.S.C. 103 as being unpatentable over Kim et al. (US 20220179661 A1) in view of Coven et al. (US 11379686 B2), further in view of Deng et al. (US 20260140728 A1).
With respect to claim 1, Kim teaches a method for deploying artificial intelligence/machine learning models as cloud-native web services, comprising:
starting, on a cloud platform, a webserver to accept requests (e.g. paragraph 0075, client 400 requesting device 100 for artificial intelligence service, and device 100 providing artificial intelligence service to the client 400; paragraph 0076, Fig. 2, electronic device 100 implemented as a server such as a cloud server);
scanning an application environment for a model loading function for an artificial intelligence/machine learning (AI/ML) model and a model invocation function for the AI/ML model (e.g. paragraph 0066, device 100 receiving artificial intelligence service function from client 300 and storing AI service function; paragraph 0068, function is an assembly of codes for executing specific operation and may include model prediction code provided by client 200; paragraph 0070, artificial intelligence service client 300 generating artificial intelligence service function for executing specific operation based on selected artificial intelligence model and model prediction code and providing it to device 100; paragraph 0071, client 200 registering trained artificial intelligence model file, prediction code, and operation information; paragraph 0072, registering trained artificial intelligence model, code, and operation information using interface provided by device 100; paragraph 0081, receiving trained AI model file, prediction code, operation information, and AI service function; paragraph 0097, when trained AI model registered, identifying container in which model is to be loaded based on attributes of containers and attribute of the registered AI model; paragraph 0101, code for loading the artificial intelligence model in the container);
configuring the webserver based on information from an AI/ML service (e.g. paragraphs 0066, 0068, 0070-0072, 0081, 0097, and 0101 as cited above; i.e. the AI service function for causing execution of a specific operation by the AI model (analogous to a model invocation function) and code/information of the model itself, including code related to/for loading the AI model (analogous to a model loading function) are registered at the device 100 which may be a server device, such that the server is configured based on the received information);
executing the model loading function to load a model object (e.g. paragraph 0101, executing code for loading the artificial intelligence model in the container);
accepting, by the webserver, an incoming AI/ML model invocation request from a client (e.g. paragraph 0075, client 400 requesting device 100 for specific artificial intelligence service; paragraph 0081, device 100 receiving request for artificial intelligence service; paragraph 0103, request for artificial intelligence service received from client 400; paragraph 0107, receiving request for face recognition service from client 400);
executing the model invocation function with data by causing the AI/ML service to execute the AI/ML model (e.g. paragraph 0075, device 100 providing artificial intelligence service to the client 400 by using artificial intelligence service function corresponding to the artificial intelligence service requested by the client; paragraph 0103, obtaining function corresponding to requested artificial intelligence service; paragraph 0106, identifying artificial intelligence model corresponding to the artificial intelligence service based on the model prediction code of the model included in the obtained function, identifying container in which the identified artificial intelligence model is loaded; paragraph 0107, obtaining face recognition function, identifying face recognition AI model corresponding to the service, and identifying the container in which the identified face recognition AI model is loaded; paragraph 0108, executing the obtained function in the container where it is identified that the artificial intelligence model for executing the obtained function is loaded); and
returning an output of the AI/ML model to the client (e.g. paragraph 0081, device 100 transmitting response to request of the artificial intelligence service; paragraph 0109-0111, obtaining result value (interference value) from input value to the trained artificial intelligence model, i.e. obtaining data for the request of the client 400 from the AI model loaded in the container, and transmitting the obtained data to the client 400).
Kim does not explicitly disclose an integration layer performing the steps above. However, Coven teaches an integration layer performing the steps (e.g. col. 7 lines 51-54, data processing technologies offered as data processing services via API; col. 8 lines 15-61, integration layer triggering execution of data processing services and receiving outputs of data processing services, which are plugged-in to the integration layer; invocation of data processing service triggered by plug-ins of the integration layer; plug-in facilitates communication between cloud based services and hosts of data processing services; plug-ins registered with the integration layer; registration of data processing service with integration layer which in turn may register existence of data processing service with cloud based collaboration platform; registration process includes registration of domain of host of data processing service, registration of mechanism to invoke the data processing service from the integration layer, registration of function of data processing service, registration of characteristics of inputs to the data processing service, registration of outputs of the data processing service, etc.; col. 11 lines 1-5, “skill” refers to a module that handles processing of content using data processing technologies; col. 20 lines 32-36, indicating that skills module may include/be implemented using machine learning models; i.e. a skill, which is a data processing technology/service, may be implemented using an AI/ML model, and may be registered for use/invocation via an integration layer which then triggers execution of the corresponding services (implemented via the model), etc. in response to user requests, etc.).
Accordingly, it would have been obvious to one of ordinary skill in the art before the effective filing date of the invention having the teachings of Kim and Coven in front of him to have modified the teachings of Kim (directed to providing artificial intelligence services), to incorporate the teachings of Coven (directed to deploying data processing service plug-ins into a cloud based collaboration platform) to include the capability to implement the method (i.e. of Kim) via an integration layer (as taught by Coven). One of ordinary skill would have been motivated to perform such a modification in order to permit integration of data processing technologies, which can provide significant value via extracted insights and are available cheaply and easily via APIs, with a cloud-based collaboration platform as described in Coven (col. 1 lines 35-53).
Kim and Coven do not explicitly disclose that the model loading function loads the AI/ML model into memory and the model invocation function defines how the AI/ML model is invoked. However, Deng teaches that the model loading function loads the AI/ML model into memory and the model invocation function defines how the AI/ML model is invoked (e.g. paragraphs 0250-0254, discussing user creating procedure.py file and calling of various functions, including at least the loadRLAlgorithm() function for loading the model.pb file which is a trained CNN model (i.e. analogous to a model loading function that loads the model into memory) and the runRLAlgorithm() function for inputting obtained data into the loaded algorithm model, executing inference using the neural network, and returning the analysis result (i.e. analogous to a model invocation function that invokes the model for execution, and therefore defines how the model is invoked); paragraph 0256, indicating that the files are compressed into a package app.tar and uploaded to the application management and control platform; paragraph 0257, indicating that the application management and control platform decompresses and obtains the files through the application parsing module; paragraph 0259, receiving files including procedure.py file and parsing the environment information required for deploying the application; paragraph 0268-0274, describing the application operation platform actually receiving the configuration information and executing the application procedure file to complete application startup and operation, including calling of the loadRLAlgorithm() function to load the model.pb file in the operation environment to obtain the trained model and calling of the runRLAlgorithm() function to input required data into the model and cause it to perform analysis and inference on the input data).
Accordingly, it would have been obvious to one of ordinary skill in the art before the effective filing date of the invention having the teachings of Kim, Coven, and Deng in front of him to have modified the teachings of Kim (directed to providing artificial intelligence services) and Coven (directed to deploying data processing service plug-ins into a cloud based collaboration platform), to incorporate the teachings of Deng (directed to a data processing method for executing a target application including an AI/ML model in an environment) to include the capability to implement the model loading function as a function which loads the AI/ML model into memory and the model invocation function as a function which defines how the AI/ML model is invoked. One of ordinary skill would have been motivated to perform such a modification in order to solve the problems of high labor cost, low efficiency, and low accuracy of information processing solutions for application deployment and release as described in Deng (paragraph 0007).
With respect to claim 8, Kim teaches a system, comprising:
a cloud platform executed by at least one computer processor and a memory(e.g. paragraph 0076, cloud server) and configured to execute an artificial intelligence/machine learning (AI/ML) service (e.g. paragraph 0063, device 100 providing artificial intelligence service; paragraph 0076, device 100 implemented as server such as a cloud server) that comprises:
an application environment (e.g. paragraph 0030, plurality of models stored in a model store; plurality of containers in which library for artificial intelligence models is loaded; paragraph 0065, artificial intelligence service providing suitable services via models performing a variety of functions; compare with specification of the instant application at paragraph 0042, application environment may include container or other deployment mechanism) that comprises a AI/ML model (e.g. paragraphs 0030 and 0065 as previously cited, plurality of models in model store, loaded into plurality of containers); and
a client electronic device comprising a client electronic device computer processor and configured to execute a client computer program (e.g. paragraph 0064, Fig. 2, client 400; paragraph 0074, client 400 corresponding to electronic device of user who desires to use AI service; paragraph 0077, client 400 implemented as electronic devices such as desktop computer, tablet PC, laptop, etc.);
wherein the system is configured to start a webserver
to accept requests (e.g. paragraph 0075, client 400 requesting device 100 for artificial intelligence service, and device 100 providing artificial intelligence service to the client 400; paragraph 0076, Fig. 2, electronic device 100 implemented as a server such as a cloud server),
to scan the application environment for a model loading function for the AI/ML model and a model invocation function for the AI/ML model (e.g. paragraph 0066, device 100 receiving artificial intelligence service function from client 300 and storing AI service function; paragraph 0068, function is an assembly of codes for executing specific operation and may include model prediction code provided by client 200; paragraph 0070, artificial intelligence service client 300 generating artificial intelligence service function for executing specific operation based on selected artificial intelligence model and model prediction code and providing it to device 100; paragraph 0071, client 200 registering trained artificial intelligence model file, prediction code, and operation information; paragraph 0072, registering trained artificial intelligence model, code, and operation information using interface provided by device 100; paragraph 0081, receiving trained AI model file, prediction code, operation information, and AI service function; paragraph 0097, when trained AI model registered, identifying container in which model is to be loaded based on attributes of containers and attribute of the registered AI model; paragraph 0101, code for loading the artificial intelligence model in the container),
to configure the webserver based on information from an AI/ML service (e.g. paragraphs 0066, 0068, 0070-0072, 0081, 0097, and 0101 as cited above; i.e. the AI service function for causing execution of a specific operation by the AI model (analogous to a model invocation function) and code/information of the model itself, including code related to/for loading the AI model (analogous to a model loading function) are registered at the device 100 which may be a server device, such that the server is configured based on the received information),
to execute the model loading function to load a model object (e.g. paragraph 0101, executing code for loading the artificial intelligence model in the container),
to accept, using the webserver, an incoming AI/ML model invocation request from a client (e.g. paragraph 0075, client 400 requesting device 100 for specific artificial intelligence service; paragraph 0081, device 100 receiving request for artificial intelligence service; paragraph 0103, request for artificial intelligence service received from client 400; paragraph 0107, receiving request for face recognition service from client 400),
to execute the model invocation function with data by causing the AI/ML service to execute the AI/ML model (e.g. paragraph 0075, device 100 providing artificial intelligence service to the client 400 by using artificial intelligence service function corresponding to the artificial intelligence service requested by the client; paragraph 0103, obtaining function corresponding to requested artificial intelligence service; paragraph 0106, identifying artificial intelligence model corresponding to the artificial intelligence service based on the model prediction code of the model included in the obtained function, identifying container in which the identified artificial intelligence model is loaded; paragraph 0107, obtaining face recognition function, identifying face recognition AI model corresponding to the service, and identifying the container in which the identified face recognition AI model is loaded; paragraph 0108, executing the obtained function in the container where it is identified that the artificial intelligence model for executing the obtained function is loaded), and
to return an output of the AI/ML model to the client (e.g. paragraph 0081, device 100 transmitting response to request of the artificial intelligence service; paragraph 0109-0111, obtaining result value (interference value) from input value to the trained artificial intelligence model, i.e. obtaining data for the request of the client 400 from the AI model loaded in the container, and transmitting the obtained data to the client 400).
Kim does not explicitly disclose an integration layer and that the integration layer performs the steps. However, Coven teaches an integration layer and that the integration layer performs the steps (e.g. col. 7 lines 51-54, data processing technologies offered as data processing services via API; col. 8 lines 15-61, integration layer triggering execution of data processing services and receiving outputs of data processing services, which are plugged-in to the integration layer; invocation of data processing service triggered by plug-ins of the integration layer; plug-in facilitates communication between cloud based services and hosts of data processing services; plug-ins registered with the integration layer; registration of data processing service with integration layer which in turn may register existence of data processing service with cloud based collaboration platform; registration process includes registration of domain of host of data processing service, registration of mechanism to invoke the data processing service from the integration layer, registration of function of data processing service, registration of characteristics of inputs to the data processing service, registration of outputs of the data processing service, etc.; col. 11 lines 1-5, “skill” refers to a module that handles processing of content using data processing technologies; col. 20 lines 32-36, indicating that skills module may include/be implemented using machine learning models; i.e. a skill, which is a data processing technology/service, may be implemented using an AI/ML model, and may be registered for use/invocation via an integration layer which then triggers execution of the corresponding services (implemented via the model), etc. in response to user requests, etc.).
Accordingly, it would have been obvious to one of ordinary skill in the art before the effective filing date of the invention having the teachings of Kim and Coven in front of him to have modified the teachings of Kim (directed to providing artificial intelligence services), to incorporate the teachings of Coven (directed to deploying data processing service plug-ins into a cloud based collaboration platform) to include the capability to implement the method (i.e. of Kim) via an integration layer (as taught by Coven). One of ordinary skill would have been motivated to perform such a modification in order to permit integration of data processing technologies, which can provide significant value via extracted insights and are available cheaply and easily via APIs, with a cloud-based collaboration platform as described in Coven (col. 1 lines 35-53).
Kim and Coven do not explicitly disclose that the model loading function loads the AI/ML model into memory and the model invocation function defines how the AI/ML model is invoked. However, Deng teaches that the model loading function loads the AI/ML model into memory and the model invocation function defines how the AI/ML model is invoked (e.g. paragraphs 0250-0254, discussing user creating procedure.py file and calling of various functions, including at least the loadRLAlgorithm() function for loading the model.pb file which is a trained CNN model (i.e. analogous to a model loading function that loads the model into memory) and the runRLAlgorithm() function for inputting obtained data into the loaded algorithm model, executing inference using the neural network, and returning the analysis result (i.e. analogous to a model invocation function that invokes the model for execution, and therefore defines how the model is invoked); paragraph 0256, indicating that the files are compressed into a package app.tar and uploaded to the application management and control platform; paragraph 0257, indicating that the application management and control platform decompresses and obtains the files through the application parsing module; paragraph 0259, receiving files including procedure.py file and parsing the environment information required for deploying the application; paragraph 0268-0274, describing the application operation platform actually receiving the configuration information and executing the application procedure file to complete application startup and operation, including calling of the loadRLAlgorithm() function to load the model.pb file in the operation environment to obtain the trained model and calling of the runRLAlgorithm() function to input required data into the model and cause it to perform analysis and inference on the input data).
Accordingly, it would have been obvious to one of ordinary skill in the art before the effective filing date of the invention having the teachings of Kim, Coven, and Deng in front of him to have modified the teachings of Kim (directed to providing artificial intelligence services) and Coven (directed to deploying data processing service plug-ins into a cloud based collaboration platform), to incorporate the teachings of Deng (directed to a data processing method for executing a target application including an AI/ML model in an environment) to include the capability to implement the model loading function as a function which loads the AI/ML model into memory and the model invocation function as a function which defines how the AI/ML model is invoked. One of ordinary skill would have been motivated to perform such a modification in order to solve the problems of high labor cost, low efficiency, and low accuracy of information processing solutions for application deployment and release as described in Deng (paragraph 0007).,
With respect to claim 15, Kim teaches a non-transitory computer readable storage medium, including instructions stored thereon, which when read and executed by one or more computer processors, cause the one or more computer processors to perform steps (e.g. paragraph 0189, described embodiments implemented as software including instructions stored in media; machine invoking instructions and operating according to the invoked instructions; paragraph 0190, instructions executed by processor; paragraph 0191, non-transitory storage medium) comprising:
starting a webserver to accept requests (e.g. paragraph 0075, client 400 requesting device 100 for artificial intelligence service, and device 100 providing artificial intelligence service to the client 400; paragraph 0076, Fig. 2, electronic device 100 implemented as a server such as a cloud server);
scanning an application environment for a model loading function for an artificial intelligence/machine learning (AI/ML) model and a model invocation function for the AI/ML model (e.g. paragraph 0066, device 100 receiving artificial intelligence service function from client 300 and storing AI service function; paragraph 0068, function is an assembly of codes for executing specific operation and may include model prediction code provided by client 200; paragraph 0070, artificial intelligence service client 300 generating artificial intelligence service function for executing specific operation based on selected artificial intelligence model and model prediction code and providing it to device 100; paragraph 0071, client 200 registering trained artificial intelligence model file, prediction code, and operation information; paragraph 0072, registering trained artificial intelligence model, code, and operation information using interface provided by device 100; paragraph 0081, receiving trained AI model file, prediction code, operation information, and AI service function; paragraph 0097, when trained AI model registered, identifying container in which model is to be loaded based on attributes of containers and attribute of the registered AI model; paragraph 0101, code for loading the artificial intelligence model in the container);
configuring the webserver based on information from an AI/ML service (e.g. paragraphs 0066, 0068, 0070-0072, 0081, 0097, and 0101 as cited above; i.e. the AI service function for causing execution of a specific operation by the AI model (analogous to a model invocation function) and code/information of the model itself, including code related to/for loading the AI model (analogous to a model loading function) are registered at the device 100 which may be a server device, such that the server is configured based on the received information);
executing the model loading function to load a model object (e.g. paragraph 0101, executing code for loading the artificial intelligence model in the container);
accepting an incoming AI/ML model invocation request from a client (e.g. paragraph 0075, client 400 requesting device 100 for specific artificial intelligence service; paragraph 0081, device 100 receiving request for artificial intelligence service; paragraph 0103, request for artificial intelligence service received from client 400; paragraph 0107, receiving request for face recognition service from client 400);
executing the model invocation function with data by causing the AI/ML service to execute the AI/ML model (e.g. paragraph 0075, device 100 providing artificial intelligence service to the client 400 by using artificial intelligence service function corresponding to the artificial intelligence service requested by the client; paragraph 0103, obtaining function corresponding to requested artificial intelligence service; paragraph 0106, identifying artificial intelligence model corresponding to the artificial intelligence service based on the model prediction code of the model included in the obtained function, identifying container in which the identified artificial intelligence model is loaded; paragraph 0107, obtaining face recognition function, identifying face recognition AI model corresponding to the service, and identifying the container in which the identified face recognition AI model is loaded; paragraph 0108, executing the obtained function in the container where it is identified that the artificial intelligence model for executing the obtained function is loaded); and
returning an output of the AI/ML model to the client (e.g. paragraph 0081, device 100 transmitting response to request of the artificial intelligence service; paragraph 0109-0111, obtaining result value (interference value) from input value to the trained artificial intelligence model, i.e. obtaining data for the request of the client 400 from the AI model loaded in the container, and transmitting the obtained data to the client 400).
Assuming arguendo that Kim does not explicitly disclose a model invocation function for the AI/ML model, Coven teaches a model invocation function for the AI/ML model (e.g. col. 7 lines 51-54, data processing technologies offered as data processing services via API; col. 8 lines 15-61, integration layer triggering execution of data processing services and receiving outputs of data processing services, which are plugged-in to the integration layer; invocation of data processing service triggered by plug-ins of the integration layer; plug-in facilitates communication between cloud based services and hosts of data processing services; plug-ins registered with the integration layer; registration of data processing service with integration layer which in turn may register existence of data processing service with cloud based collaboration platform; registration process includes registration of domain of host of data processing service, registration of mechanism to invoke the data processing service from the integration layer, registration of function of data processing service, registration of characteristics of inputs to the data processing service, registration of outputs of the data processing service, etc.; col. 11 lines 1-5, “skill” refers to a module that handles processing of content using data processing technologies; col. 20 lines 32-36, indicating that skills module may include/be implemented using machine learning models; i.e. a skill, which is a data processing technology/service, may be implemented using an AI/ML model, and corresponding functions for invoking the skill/model may be registered).
Accordingly, it would have been obvious to one of ordinary skill in the art before the effective filing date of the invention having the teachings of Kim and Coven in front of him to have modified the teachings of Kim (directed to providing artificial intelligence services), to incorporate the teachings of Coven (directed to deploying data processing service plug-ins into a cloud based collaboration platform) to include the capability to scan/register a function/mechanism to invoke the AI/ML model (as taught by Coven). One of ordinary skill would have been motivated to perform such a modification in order to permit integration of data processing technologies, which can provide significant value via extracted insights and are available cheaply and easily via APIs, with a cloud-based collaboration platform as described in Coven (col. 1 lines 35-53).
Kim and Coven do not explicitly disclose that the model loading function loads the AI/ML model into memory and the model invocation function defines how the AI/ML model is invoked. However, Deng teaches that the model loading function loads the AI/ML model into memory and the model invocation function defines how the AI/ML model is invoked (e.g. paragraphs 0250-0254, discussing user creating procedure.py file and calling of various functions, including at least the loadRLAlgorithm() function for loading the model.pb file which is a trained CNN model (i.e. analogous to a model loading function that loads the model into memory) and the runRLAlgorithm() function for inputting obtained data into the loaded algorithm model, executing inference using the neural network, and returning the analysis result (i.e. analogous to a model invocation function that invokes the model for execution, and therefore defines how the model is invoked); paragraph 0256, indicating that the files are compressed into a package app.tar and uploaded to the application management and control platform; paragraph 0257, indicating that the application management and control platform decompresses and obtains the files through the application parsing module; paragraph 0259, receiving files including procedure.py file and parsing the environment information required for deploying the application; paragraph 0268-0274, describing the application operation platform actually receiving the configuration information and executing the application procedure file to complete application startup and operation, including calling of the loadRLAlgorithm() function to load the model.pb file in the operation environment to obtain the trained model and calling of the runRLAlgorithm() function to input required data into the model and cause it to perform analysis and inference on the input data).
Accordingly, it would have been obvious to one of ordinary skill in the art before the effective filing date of the invention having the teachings of Kim, Coven, and Deng in front of him to have modified the teachings of Kim (directed to providing artificial intelligence services) and Coven (directed to deploying data processing service plug-ins into a cloud based collaboration platform), to incorporate the teachings of Deng (directed to a data processing method for executing a target application including an AI/ML model in an environment) to include the capability to implement the model loading function as a function which loads the AI/ML model into memory and the model invocation function as a function which defines how the AI/ML model is invoked. One of ordinary skill would have been motivated to perform such a modification in order to solve the problems of high labor cost, low efficiency, and low accuracy of information processing solutions for application deployment and release as described in Deng (paragraph 0007).,
With respect to claims 5, 12, and 19, Kim in view of Coven, further in view of Deng teaches all of the limitations of claims 1, 8, and 15 as previously discussed, and Coven further teaches the method further comprising: transforming, by the integration layer/wherein the integration layer is configured to transform/transforming, a data structure from the incoming AI/ML model invocation request to a format for the AI/ML model; and transforming/to transform, by the integration layer, the output of the AI/ML model for transport to the client (e.g. col. 7 lines 51-54, data processing technologies offered as data processing services via API; col. 8 lines 28-31, invocation of data processing service to perform particular function on particular content item triggered by plug-in of integration layer; col. 11 lines 1-5, “skill” refers to a module that handles processing of content using data processing technologies; col. 20 lines 32-36, indicating that skills module may include/be implemented using machine learning models; col. 34 line 36-col. 35 line 30, Fig. 11B, executing skill upon receiving trigger; using common input format 1180 to invoke skills at different providers; determining which services apply to skills processing that underlies the received trigger; accessing repository of metadata format definitions comprising formats that can be used when interacting with any known provider, and that apply to data content and formatting when sending a request to a service provider as well as formats that apply to data content and formatting when processing data items that are received from a service provider in response to a request; selecting applicable metadata definitions that apply to the determined services, and sending request to external data processing services; after selecting applicable metadata definition pertaining to request, metadata is populated to external data processing services using the selected common input format; data processing services then process the request and provide data that corresponds to the request; outputs from various data processing services differ; accordingly, outputs undergo process to bring output data from any particular data processing service into common format used by downstream processing, as is also shown in Fig. 11C; i.e. the skill (implemented via ML model) is invoked via the integration layer, where this includes selecting an applicable format to transform the trigger/invocation/request into in order to communicate it to the applicable data processing service, and further includes receiving corresponding response data from the data processing service and transforming it into a corresponding format for use in downstream processing).
Accordingly, it would have been obvious to one of ordinary skill in the art before the effective filing date of the invention having the teachings of Kim, Coven, and Deng in front of him to have modified the teachings of Kim (directed to providing artificial intelligence services) and Deng (directed to a data processing method for executing a target application including an AI/ML model in an environment), to incorporate the teachings of Coven (directed to deploying data processing service plug-ins into a cloud based collaboration platform) to include the capability to implement the method (i.e. of Kim) via an integration layer, where the integration layer transforms the data format of the invocation request into a format suitable for transmission to the ML model (i.e. such as a skill/data processing service implemented using an ML model as taught by Coven) and further transforms output data received from the ML model into a format suitable for downstream processing (as taught by Coven, where the downstream processing may include providing the response to the client as taught by Kim). One of ordinary skill would have been motivated to perform such a modification in order to permit integration of data processing technologies, which can provide significant value via extracted insights and are available cheaply and easily via APIs, with a cloud-based collaboration platform as described in Coven (col. 1 lines 35-53).
With respect to claims 6 and 13, Kim in view of Coven, further in view of Deng teaches all of the limitations of claims 1 and 8 as previously discussed, and Kim further teaches wherein executing the model invocation function passes the data and the model object to AI/ML service (e.g. paragraph 0155, receiving request from client 400 including input data for AI service together with information on the model for providing the AI service; if requesting face recognition service, receiving an image file as input data together with the request for the service from the client; paragraph 0162, invoker transferring code of the function and the input data to the container in which the corresponding AI model is loaded; i.e. the request/function to invoke the model includes at least input data to be processed by the model and identification of the model itself).
With respect to claims 7, 14, and 20, Kim in view of Coven, further in view of Deng teaches all of the limitations of claims 1, 8, and 15 as previously discussed, and Kim further teaches receiving, from the AI/ML service/wherein the system is configured to receive, a health status (e.g. paragraphs 0119-0124, container executing function in first state, and determining the container is then in the second state; first state indicates state where artificial intelligence model is not loaded in the container or a pre-warm state where the model is loaded in the container and the container does not execute the function; second state may refer to a warm state where the container executes the function using the loaded model; if number of containers in first state is less than predetermined number, generating new container; maintaining predetermined number of containers in first state; determining numbers of containers in different states (first and second states); if container in second state does not receive request for AI service using the model during predetermined time, killing the container; i.e. the system continually receives/monitors status information regarding the current number of containers and their respective operating states, including at least a pre-warm state where the container is available and not executing, and a warm state where the container is currently executing, analogous to a health status; compare with specification of the instant application at paragraphs 0053-0054, indicating that the “health status” provides an indication regarding whether a model is available or unavailable, where an affirmative response indicates that the model object is loaded and available in memory).
Coven teaches performing the steps using the integration layer (e.g. col. 7 lines 51-54, data processing technologies offered as data processing services via API; col. 8 lines 15-61, integration layer triggering execution of data processing services and receiving outputs of data processing services, which are plugged-in to the integration layer; invocation of data processing service triggered by plug-ins of the integration layer; plug-in facilitates communication between cloud based services and hosts of data processing services; plug-ins registered with the integration layer; registration of data processing service with integration layer which in turn may register existence of data processing service with cloud based collaboration platform; registration process includes registration of domain of host of data processing service, registration of mechanism to invoke the data processing service from the integration layer, registration of function of data processing service, registration of characteristics of inputs to the data processing service, registration of outputs of the data processing service, etc.; col. 11 lines 1-5, “skill” refers to a module that handles processing of content using data processing technologies; col. 20 lines 32-36, indicating that skills module may include/be implemented using machine learning models; i.e. a skill, which is a data processing technology/service, may be implemented using an AI/ML model, and may be registered for use/invocation via an integration layer which then triggers execution of the corresponding services (implemented via the model), etc. in response to user requests, etc.).
Accordingly, it would have been obvious to one of ordinary skill in the art before the effective filing date of the invention having the teachings of Kim, Coven, and Deng in front of him to have modified the teachings of Kim (directed to providing artificial intelligence services) and Deng (directed to a data processing method for executing a target application including an AI/ML model in an environment), to incorporate the teachings of Coven (directed to deploying data processing service plug-ins into a cloud based collaboration platform) to include the capability to implement the method (i.e. of Kim) via an integration layer (as taught by Coven). One of ordinary skill would have been motivated to perform such a modification in order to permit integration of data processing technologies, which can provide significant value via extracted insights and are available cheaply and easily via APIs, with a cloud-based collaboration platform as described in Coven (col. 1 lines 35-53).
Claims 2, 4, 9, 11, 16, and 18 are rejected under 35 U.S.C. 103 as being unpatentable over Kim in view of Coven, further in view of Deng, further in view of Nikitin et al. (US 20220391748 A1).
With respect to claims 2, 9, and 16, Kim in view of Coven, further in view of Deng teaches all of the limitations of claims 1, 8, and 15 as previously discussed. Kim and Coven do not explicitly disclose wherein the webserver is an HTTP webserver. However, Nikitin teaches wherein the webserver is an HTTP webserver (e.g. paragraph 0012, containers for ML models implemented with HTTP web service and dependencies incorporated into the container; paragraph 0044, server 207 servicing call from client 203, causing calling of model load function, etc.; paragraph 0056, ML serving infrastructure implemented in server devices with software to implement the ML serving infrastructure; paragraph 0060, system 440 including ML serving infrastructure; paragraph 0066, system 440 including server devices to handle user requests; paragraph 0069, user device 480 accessing data and applications hosted by system 440, such as using HTTP; sending and receiving HTTP messages to and from servers of system 440).
Accordingly, it would have been obvious to one of ordinary skill in the art before the effective filing date of the invention having the teachings of Kim, Coven, Deng, and Nikitin in front of him to have modified the teachings of Kim (directed to providing artificial intelligence services), Deng (directed to a data processing method for executing a target application including an AI/ML model in an environment), and Coven (directed to deploying data processing service plug-ins into a cloud based collaboration platform), to incorporate the teachings of Nikitin (directed to API based container service for supporting machine learning applications) to include the capability to implement the server (i.e. of Kim) as an HTTP webserver (as taught by Nikitin). One of ordinary skill would have been motivated to perform such a modification in order to enable ML model developers to customize serving containers for models such that the developers are assured that the container can be successfully deployed to ML serving infrastructure without significant investment of development time and manpower as described in Nikitin (paragraph 0012).
With respect to claims 4, 11, and 18, Kim in view of Coven, further in view of Deng teaches all of the limitations of claims 1, 8, and 15 as previously discussed. Kim and Coven do not explicitly disclose wherein the incoming AI/ML model invocation request is received at a http address. However, Nikitin teaches wherein the incoming AI/ML model invocation request is received at a http address (e.g. paragraph 0012, containers for ML models implemented with HTTP web service and dependencies incorporated into the container; paragraph 0044, server 207 servicing call from client 203, causing calling of model load function, etc.; paragraph 0056, ML serving infrastructure implemented in server devices with software to implement the ML serving infrastructure; paragraph 0060, system 440 including ML serving infrastructure; paragraph 0066, system 440 including server devices to handle user requests; paragraph 0069, user device 480 accessing data and applications hosted by system 440, such as using HTTP; sending and receiving HTTP messages to and from servers of system 440; i.e. where the request is received at a server via HTTP, this will include receipt of the request at an HTTP address).
Accordingly, it would have been obvious to one of ordinary skill in the art before the effective filing date of the invention having the teachings of Kim, Coven, Deng, and Nikitin in front of him to have modified the teachings of Kim (directed to providing artificial intelligence services), Deng (directed to a data processing method for executing a target application including an AI/ML model in an environment), and Coven (directed to deploying data processing service plug-ins into a cloud based collaboration platform), to incorporate the teachings of Nikitin (directed to API based container service for supporting machine learning applications) to include the capability to implement the server (i.e. of Kim) as an HTTP webserver, such that the invocation requestion is received at an HTTP address (as taught by Nikitin). One of ordinary skill would have been motivated to perform such a modification in order to enable ML model developers to customize serving containers for models such that the developers are assured that the container can be successfully deployed to ML serving infrastructure without significant investment of development time and manpower as described in Nikitin (paragraph 0012).
Claims 3, 10, and 17 are rejected under 35 U.S.C. 103 as being unpatentable over Kim in view of Coven, further in view of Deng, further in view of Liu et al. (US 20240086727 A1).
With respect to claims 3, 10, and 17, Kim in view of Coven, further in view of Deng teaches all of the limitations of claims 1, 8, and 15 as previously discussed. Kim and Coven do not explicitly disclose wherein the integration layer is configured to configure the webserver/wherein the webserver is configured by specifying a number of invocations to handle concurrently and/or a number of central processing units available. However, Liu teaches wherein the integration layer configures the webserver/wherein the webserver is configured by specifying a number of invocations to handle concurrently and/or a number of central processing units available (e.g. paragraph 0015, historical machine learning model training environment resource properties, such as number of nodes, number of CPUs, etc.; automatically making recommended adjustments by changing environment runtime size, such as number or amount of resources; paragraph 0045, machine learning model training environment runtime properties include pod CPU core number, pod memory size, cluster controller CPU core number, cluster controller memory size, cluster worker node CPU core number, cluster worker node memory size, cluster worker node number, etc.; paragraph 0047, automatically building efficient machine learning model training runtime in cloud environment).
Accordingly, it would have been obvious to one of ordinary skill in the art before the effective filing date of the invention having the teachings of Kim, Coven, Deng, and Liu in front of him to have modified the teachings of Kim (directed to providing artificial intelligence services), Deng (directed to a data processing method for executing a target application including an AI/ML model in an environment), and Coven (directed to deploying data processing service plug-ins into a cloud based collaboration platform), to incorporate the teachings of Liu (directed to automatically building efficient machine learning model training environments) to include the capability to automatically adjust the runtime environment for the model (i.e. by configuring the webserver of Kim, where this may be performed by an integration layer as taught by Coven) by specifying a number of CPUs available (as taught by Liu). One of ordinary skill would have been motivated to perform such a modification in order to automatically build an efficient machine learning model runtime in a cloud environment, to decrease time, resource, and cost consumption for the user and the cloud environment, increasing efficiency and performance of the machine learning environments as described in Liu (paragraph 0047).
It is noted that any citation to specific pages, columns, lines, or figures in the prior art references and any interpretation of the references should not be considered to be limiting in any way. “The use of patents as references is not limited to what the patentees describe as their own inventions or to the problems with which they are concerned. They are part of the literature of the art, relevant for all they contain,” In re Heck, 699 F.2d 1331, 1332-33, 216 USPQ 1038, 1039 (Fed. Cir. 1983) (quoting in re Lemelson, 397 F.2d 1006, 1009, 158 USPQ 275, 277 (GCPA 1968)). Further, a reference may be relied upon for all that it would have reasonably suggested to one having ordinary skill the art, including nonpreferred embodiments. Merck & Co, v. Biocraft Laboratories, 874 F.2d 804, 10 USPQ2d 1843 (Fed. Cir.), cert, denied, 493 U.S. 975 (1989). See also Upsher-Smith Labs. v. Pamlab, LLC, 412 F,3d 1319, 1323, 75 USPQ2d 1213, 1215 (Fed. Cir, 2005): Celeritas Technologies Ltd. v. Rockwell International Corp., 150 F.3d 1354, 1361, 47 USPQ2d 1516, 1522-23 (Fed. Cir. 1998).
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant’s disclosure.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to JEREMY L STANLEY whose telephone number is (469)295-9105. The examiner can normally be reached on Monday-Friday from 9:00 AM to 5:00 PM CST.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Abdullah Al Kawsar, can be reached at telephone number (571) 270-3169. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
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.
Information regarding the status of an application may be obtained from Patent Center and the Private Patent Application Information Retrieval (PAIR) system. Status information for published applications may be obtained from Patent Center or Private PAIR. Status information for unpublished applications is available through Patent Center and Private PAIR for authorized users only. Should you have questions about access to the Private PAIR system, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free).
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) Form at https://www.uspto.gov/patents/uspto-automated- interview-request-air-form.
/JEREMY L STANLEY/
Primary Examiner, Art Unit 2127
Prosecution Insights