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 Amendment
This office action is final and in response to the amendment filed on July 21, 2026, which was in response to the non-final office action mailed April 21, 2026.
Appropriate corrections have been made to the claim objections. Objection is withdrawn.
Claims 1, 3, 20, 21, and 22 have been amended.
Claims 2 and 15 have been cancelled.
The remaining Claims 1, 3-14, and 16-24 are pending, have been examined, and are rejected.
Information Disclosure Statement
Acknowledgment is made of the information disclosure statements filed May 23, 2023, which comply with 37 CFR 1.97. As such, the information disclosure statements have been placed in the application file and the information referred to therein has been considered by the examiner.
Priority
Applicant’s claim for the benefit of a prior-filed application under 35 U.S.C. 119(e) or under 35 U.S.C. 120, 121, 365(c), or 386(c) is acknowledged. The present application claims priority to U.S. Provisional application No. 63/345,232 filed on May 24, 2022.
Applicant has not complied with one or more conditions for receiving the benefit of an earlier filing date under 35 U.S.C. 119(e) as follows:
The later-filed application must be an application for a patent for an invention which is also disclosed in the prior application (the parent or original non-provisional application or provisional application). The disclosure of the invention in the parent application and in the later-filed application must be sufficient to comply with the requirements of 35 U.S.C. 112(a) or the first paragraph of pre-AIA 35 U.S.C. 112, except for the best mode requirement. See Transco Products, Inc. v. Performance Contracting, Inc., 38 F.3d 551, 32 USPQ2d 1077 (Fed. Cir. 1994).
The disclosure of the prior-filed application, U.S. Provisional application No. 63/345,232 (hereinafter “the ‘232 provisional application”) fails to provide adequate support or enablement in the manner provided by 35 U.S.C. 112(a) or pre-AIA 35 U.S.C. 112, first paragraph for one or more claims of this application.
Dependent claims 13 and 20 recite, using respective similar language, “wherein the one or more selected computing models include a large language model”, Claim 23 recites, using respective similar language, “wherein the first computing model is a first large language model.”, and Claim 24 recites, using respective similar language, “wherein the second computing model is a second large language model.”
The as-filed original specification of the ‘232 provisional application fails to provide adequate support or enablement for at least these elements of claims 13, 20, 23, and 24.
Therefore, the effective filing date for claims 13, 20, 23, and 24 of the instant application is the effective filing date of the instant, non-provisional application, May 23, 2023.
Each claim will receive benefit of the earliest filing date above for which a continuous chain of support can be established for the entirety of the claim.
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.
This application currently names joint inventors. In considering patentability of the claims the examiner presumes that the subject matter of the various claims was commonly owned as of the effective filing date of the claimed invention(s) absent any evidence to the contrary. Applicant is advised of the obligation under 37 CFR 1.56 to point out the inventor and effective filing dates of each claim that was not commonly owned as of the effective filing date of the later invention in order for the examiner to consider the applicability of 35 U.S.C. 102(b)(2)(C) for any potential 35 U.S.C. 102(a)(2) prior art against the later invention.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or non-obviousness.
Claims 1, 3, 4, 9, 10, 14, 16, 21, and 22 under 35 U.S.C. 103 as being unpatentable over Khare et al. (US20220129334A1, filed on January 10, 2022, hereinafter “Khare”), in view of Deshpande et al. (US20190279114A1, filed on March 8, 2018, hereinafter “Deshpande”). Both references were effectively filed before the effective filing date of this application, i.e., May 23, 2023, as well as the provisional filing date of May 24, 2022, where it applies.
With respect to Claims 1 and 14:
Khare teaches:
"A method for model selection and update operation, the method comprising: receiving a model request including one or more request parameters, the one or more request parameters including at least one selected from a group consisting of: a type of computing model, a processing characteristic, and a data characteristic;" (Paragraph [0078] discloses that a query request is received, “a request is received for a listed repository item (algorithm/model/data/pipeline/notebook) suggestion. For example, a query such as that detailed with respect to FIG.2 is received” Paragraph [0036] further discloses the specific parameters associated with this request, “A query may include one or more of: a category (and subcategory), data information (format of what is available to the requester), resource availability (what resources are available for use such as a processor type), timing information (such as desired latency), an indication of a pipeline that the model/algorithm/data is to be used for), accuracy desired, type of content desired (algorithm, model, and/or data), etc.”) Examiners Note: As per the reference a, “type of content desired” can take the place of “type of computing model”, “resource availability” can take the place of “processing characteristic”, and “data information” can take the place of “data characteristic” in regards to this disclosure.
“receiving information associated with a plurality of computing models from a model repository;” (Paragraph [0023] teaches that a model repository can be accessed to retrieve information associated with various models, “The web services model repository service 121 allows for a producer to share generated content with others. The content that is shared is searchable as listings. In particular, a requester interfaces with a model/algorithm/data application programming interface (API) frontend 109 to find and select shared content through requests serviced by a publishing/listing agent 125.” Paragraph [0019] further teaches that accessibility to information related to multiple variants of computing models can be achieved, “producers who produce data, algorithms, and/or models make them available to third parties via a registry-based system. Requesters search this register to find algorithms, models, and/or data for their own purposes. In some embodiments, one or more schemas are utilized to build machine learning pipelines using one or more of the algorithms, models, and/or data made available via the registry-based system.”)
“selecting one or more computing models based upon the model request;” (Paragraph [0078] discloses determining and selecting a specified computing model based on the model request, “At 801, in some embodiments, a request is received for a listed repository item (algorithm/model/data/pipeline/notebook) suggestion. For example, a query such as that detailed with respect to FIG.2 is received.” Paragraph [0079] describes the selecting of a computing model based on the model request (query), “Using the details of the query, one or more listed repository items that may meet the desires of the query is determined and provided to the requester at 803.”)
“compiling a container request based on the model request and the one or more selected computing models,...” (Paragraph [0070] discloses compiling a container request, “A request to list a container is received at 701. For example, a producer requests that a container in the model/algorithm container registry 105 or package published in the source control service 107 is to be listed by the publishing listing agent 125 as available for third party use.” Paragraph [0082] mentions that a container request incorporates the selected computing model that was previously determined by the model request, “In some embodiments, the selected algorithm or model is a part of a container and a copy of the container is allocated to be executed.” Paragraph [0065] discusses the presence of processors/hardware that can perform this container request process, “the publishing/listing agent 125 is provided with hardware configurations to use, or encryption to adhere to, etc.”)
“…the container request including a request using a container application programming interface (API), the container API being defined by a container infrastructure;” (Paragraph [0032] discloses that the request is sent through an API, “an API gateway 103 takes in a request at circle 8 to start execution (e.g., inference) or training by an execution service 111. The API gateway 103 instructs the execution service 111 at circle 9A.” Paragraph [0074] further discloses that the containerized model is invoked through an API, “The microservice is wrapped with an API at 711. For example, input/output definitions are created such that an API call can be made.”)
“transmitting the container request to a container infrastructure;” (Paragraph [0068] discloses the initial transmission of the container request after it’s been fulfilled, “When the verification was successful, the package is published in the model/algorithm container registry 105. Published containers are available to the publishing/listing agent 125 to be served as a potential result to a code or data query”. Paragraph [0082] discloses pipeline infrastructure (container infrastructure) that takes in all container request related information in order to perform a particular task, “The selected content may be a part of a pipeline, and, in those cases, resources are allocated for the pipeline if it is ready for execution (for example, contains only models). For example, a request to perform one or more of these acts is received by the model/algorithm/data API frontend 109 which calls on the publishing/listing agent 125 to provide necessary information to an execution service 111 (such as a location of the selected algorithm, model, data, pipeline, and/or notebook in algorithm/model/data store 123) which then allocates execution resources 113 including compute resources 117 and storage 115.”) Examiner’s Note: Execution service 111, execution resources 113, and container registry 105 together constitute the container infrastructure.
“causing the container infrastructure to allocate one or more resources for running the one or more selected computing models;” (Paragraph [0082] discloses, “Once the requester has the algorithm, model, data, notebook, and/or pipeline it requested, resources are allocated (per a request from the request) to train the selected suggested algorithm(s), execute the selected model(s), or use selected data in training an algorithm,” and the execution service “then allocates execution resources 113 including compute resources 117 and storage 115.”)
Khare does not appear to explicitly disclose:
“…the container API being defined by a container infrastructure;”
However, Deshpande teaches:
“…the container API being defined by a container infrastructure;” (Paragraph [0038] discloses a container infrastructure that defines its own API and receives provisioning requests through it, “Backend computing environment 104 may further include a container manager 110. Container manager 110 may provide an application programming interface (API) endpoint for host 108. For example, container manager 110 may be accessible to users via a REST (representational state transfer) API (application programming interface) or any other client-server communication protocol…Container manager 110 may receive provisioning requests submitted (e.g., via REST API calls) and provision a container 114 based on the provisioning requests.” Paragraph [0045] further discloses that the infrastructure’s API includes APIs dedicated to resource allocation, “Backend computing environment 104 may further include a set of APIs 204 having one or more workflows 208…For example, workflow 208 may include an allocation phase and APIs 204 may include a given subset of APIs corresponding to the allocation phase.” Paragraph [0057] further discloses that a request for a selected machine learning model is submitted through that API, “client device 102 may submit a provisioning request via REST API calls to container manager 110 for the provisioning of one or more containers 114,” where “the one or more parameters includes at least a machine learning model type corresponding to the machine learning model requested for analysis.” Paragraph [0065] discloses that the container infrastructure then provisions the container having the selected model, “container manager 110 provisions container 114 using container image 115 having machine learning model 224.”)
Khare and Deshpande are analogous art and in the same field of invention because both references pertain to provisioning and executing containerized machine learning models that are selected in response to a requester’s request. While Khare teaches compiling a request from a model query and the selected model and sending that request through API gateway 103 to an execution service that allocates resources and executes the selected model’s container, Deshpande teaches submitting such a request through an API endpoint defined by the container manager of the container infrastructure. 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 implement the base reference of Khare (selecting machine learning models from a repository and executing their containers on allocated resources) with the teachings of Deshpande (provisioning machine learning model containers through the container manager’s own API) in order to standardize the manner in which containers for selected models are requested and provisioned. Using the native interface of the container infrastructure is a known technique that yields the predictable result of provisioning the selected model’s container. One of ordinary skill in the art would be motivated to do so because by integrating Deshpande’s framework into the methods of Khare one would be able to “allow end users to quickly prepare a container image and use the container image to provision one or more containers” {[0034] of Deshpande}.
Therefore, Claims 1 and 14 are rejected.
With respect to Claim 3 which is dependent on Claim 1:
Khare-Deshpande teaches:
“The method of claim 1, wherein the one or more resources includes one or more computing resources or one or more storage resources.” (Khare Paragraph [0082] teaches the inclusion of computing and storage resources, “…which then allocates execution resources 113 including compute resources 117 and storage 115.”)
Therefore, Claim 3 is rejected.
With respect to Claims 4 and 16:
Khare-Deshpande teaches:
“The method of claim 1, wherein the compiling a container request based on the model request and the one or more selected computing models comprises: extracting one or more model parameters from the model request;” (Khare Paragraph [0030] discloses accessing the model request to extract relevant components based on accessibility levels, “In some embodiments, the user account 113 for the requester is accessed to see what the requester is allowed to access in terms of hardware, etc.” Khare Paragraph [0031] further showcases extracting the model parameters from the model request that fit the characteristics of the request, “The requester selects content from the result and provides this selection and/or code using this selection at circle 6. In some embodiments, interaction also includes a request to execute or train. In some embodiments, the publishing/listing agent 125 selects what it feels is best…”)
“and generating the container request based upon the one or more model parameters; wherein the container request is generated in a format compliant with an interface of the container infrastructure.” (Khare Paragraph [0025] discloses the architecture and structure of the container request with respect to the container infrastructure, “In some embodiments, container images include one or more layers, where each layer represents executable instructions. Some or all of the executable instructions together represent an algorithm that defines a machine learning model. The executable instructions (e.g., the algorithm) can be written in any programming language (e.g., Python, Ruby, C++, Java, etc.). Khare Paragraph [0036] further clarifies that formatting is taken into consideration when reviewing a potential model request (query) for a container request, “A request query comes into the publishing/listing agent 125 into a query evaluation service (or engine) 219. A query may include one or more of: a category (and subcategory), data information (format of what is available to the requester),”)
Therefore, Claims 4 and 16 are rejected.
With respect to Claim 9:
Khare-Deshpande teaches:
“The method of claim 1, further comprising: compiling a processing pipeline based at least in part upon the model request.” (Khare Paragraph [0023] discloses the presence of a model request, “In particular, a requester interfaces with a model/ algorithm/data application programming interface (API) frontend 109 to find and select shared content through requests serviced by a publishing/listing agent 125.” Khare Paragraph [0025] further discloses compiling a processing pipeline based on that model request, “the publishing listing agent 125 is used by a requester to build a pipeline and/or cause execution or training of a selected model or algorithm”)
Therefore, Claim 9 is rejected.
With respect to Claim 10:
Khare-Deshpande teaches:
“The method of claim 9, wherein the processing pipeline includes at least a first computing model and a second computing model operating sequentially, such that an output of the first computing model is an input of the second computing model.” (Khare Paragraph [0089] references a processing pipeline that includes a first model and a second model operating sequentially on it to where the output of the first model is used as input for the second model, “The post-training pipeline 921 includes a first model (model 1) 923 that takes in data of format X and outputs data in format Y. The output (Y) of the model 1 903 is an input into model 2 925.”
Therefore, Claim 10 is rejected.
With respect to Claim 21:
Khare teaches:
"A method for model selection and update operation, the method comprising: receiving a first model request including one or more first request parameters, the one or more first request parameters including at least one selected from a group consisting of: a type of computing model, a processing characteristic, and a data characteristic;" (Paragraph [0078] discloses that a query request for a first model is received, “a request is received for a listed repository item (algorithm/model/data/pipeline/notebook) suggestion. For example, a query such as that detailed with respect to FIG.2 is received” Paragraph [0036] further discloses the specific parameters associated with this request, “A query may include one or more of: a category (and subcategory), data information (format of what is available to the requester), resource availability (what resources are available for use such as a processor type), timing information (such as desired latency), an indication of a pipeline that the model/algorithm/data is to be used for), accuracy desired, type of content desired (algorithm, model, and/or data), etc.”) Examiners Note: As per the reference a, “type of content desired” can take the place of “type of computing model”, “resource availability” can take the place of “processing characteristic”, and “data information” can take the place of “data characteristic” in regards to this disclosure. See claim 1, [0023], and [0019].
“receiving information associated with a plurality of computing models from a model repository;” (Paragraph [0023] teaches that a model repository can be accessed to retrieve information associated with various models, “The web services model repository service 121 allows for a producer to share generated content with others. The content that is shared is searchable as listings. In particular, a requester interfaces with a model/algorithm/data application programming interface (API) frontend 109 to find and select shared content through requests serviced by a publishing/listing agent 125.” Paragraph [0019] further teaches that accessibility to information related to multiple variants of computing models can be achieved, “producers who produce data, algorithms, and/or models make them available to third parties via a registry-based system. Requesters search this register to find algorithms, models, and/or data for their own purposes. In some embodiments, one or more schemas are utilized to build machine learning pipelines using one or more of the algorithms, models, and/or data made available via the registry-based system.”)
“selecting at least a first computing model and a second computing model based upon the first model request;” (Paragraph [0078] discloses determining and selecting a specified first computing model based on the model request, “At 801 , in some embodiments, a request is received for a listed repository item (algorithm/model/data/pipeline/notebook) suggestion. For example, a query such as that detailed with respect to FIG.2 is received.” Paragraph [0079] describes the selecting of a first computing model based on the model request (query), “Using the details of the query, one or more listed repository items that may meet the desires of the query is determined and provided to the requester at 803.”)
“compiling a processing pipeline based at least in part upon the first model request, the processing pipeline including at least the first computing model and the second computing model;” (Paragraph [0082] discloses compiling a processing pipeline based on a first model request for a first model, “For example, when the publishing/listing agent 125 is helping a user build a pipeline to perform a task (or tasks) this type of request occurs. The publishing/listing agent 125 will evaluate the pipeline as it exists and make the necessary connections within the pipeline at 809.” Paragraph [0089] further teaches that this processing pipeline exists to handle the training of a first and second computing model, “The post-training pipeline 921 includes a first model (model 1) 923 that takes in data of format X and outputs data in format Y. The output (Y) of the model 1 903 is an input into model 2 925.”)
“compiling a first container request based on the first model request, the first computing model, and the second computing model, the first container request including a request using a container application programming interface (API), the container API being defined by a container infrastructure;” (Paragraph [0070] discloses compiling a first container request, “A request to list a container is received at 701. For example, a producer requests that a container in the model/algorithm container registry 105 or package published in the source control service 107 is to be listed by the publishing listing agent 125 as available for third party use.” Paragraph [0082] mentions that a container request incorporates the selected computing model(s) that were previously determined by the model request(s), “In some embodiments, the selected algorithm or model is a part of a container and a copy of the container is allocated to be executed.” Paragraph [0032] discloses that “the publishing/listing agent 125…fills in a provided pipeline with the selection prior to the instruction to the execution service 111 to execute the pipeline,” and that “an API gateway 103 takes in a request …to start execution.” Paragraph [0032] discloses that a single request is compiled from the pipeline containing both selected models and is sent through an API, “the publishing/listing agent 125 acts as an intermediary between the frontend 109 and the API gateway 103 and fills in a provided pipeline with the selection prior to the instruction to the execution service 111 to execute the pipeline or training of components thereof,” where “an API gateway 103 takes in a request at circle 8 to start execution (e.g., inference) or training by an execution service 111.” Paragraph [0082] further discloses that the request for the pipeline carries the information of the selected models, “The selected content may be a part of a pipeline and, in those cases, resources are allocated for the pipeline if it is ready for execution (for example, contains only models). For example, a request to perform one or more of these acts is received by the model/algorithm/data API frontend 109 which calls on the publishing/listing agent 125 to provide necessary information to an execution service 111 (such as a location of the selected algorithm, model, data, pipeline, and/or notebook in algorithm/model/data store 123),” and “the selected algorithm or model is a part of a container and a copy of the container is allocated to be executed.”) Examiner’s Note: Because agent 125 fills the pipeline containing both the first and second computing models before the single instruction to execute the pipeline is issued, the resulting request is compiled based on the first model request, the first computing model, and the second computing model.) See claim 1 for the remainder of these limitations, including the teachings of Deshpande regarding the container API defined by the container infrastructure.
“transmitting the first container request to a container infrastructure;” (Paragraph [0068] discloses the initial transmission of the first container request after it’s been fulfilled, “When the verification was successful, the package is published in the model/algorithm container registry 105. Published containers are available to the publishing/listing agent 125 to be served as a potential result to a code or data query”. Paragraph [0082] discloses pipeline infrastructure (container infrastructure) that takes in all container request related information in order to perform a particular task, “The selected content may be a part of a pipeline and, in those cases, resources are allocated for the pipeline if it is ready for execution (for example, contains only models). For example, a request to perform one or more of these acts is received by the model/algorithm/data API frontend 109 which calls on the publishing/listing agent 125 to provide necessary information to an execution service 111 (such as a location of the selected algorithm, model, data, pipeline, and/or notebook in algorithm/model/data store 123) which then allocates execution resources 113 including compute resources 117 and storage 115.”)
“causing the container infrastructure to allocate one or more resources for running the first computing model and the second computing model; wherein the method is performed using one or more processors.” (Paragraph [0036] reinforces that this method can be performed via processors as per the initial first model request (query), “A query may include one or more of: a category (and subcategory), data information (format of what is available to the requester), resource availability (what resources are available for use such as a processor type)…” Paragraph [0083] discloses, “In some embodiments, different resources are allocated for different stages of the pipeline.” See also Paragraphs [0082] and [0020]. Paragraph [0082] discloses that “resources are allocated for the pipeline if it is ready for execution (for example, contains only models)” and that execution service 111 “allocates execution resources 113 including compute resources 117 and storage 115.” Paragraph [0083] further discloses allocating resources for each model of the pipeline, “In some embodiments, different resources are allocated for different stages of the pipeline. These different resources are selected based on the information of the listing (such as suggested resource requirements, latency, etc.).”) Examiner’s Note: Model 1 (923) and Model 2 (925) occupy different stages of post-training pipeline 921 [0089], so allocating resources for “different stages of the pipeline” is allocating resources for running the first computing model and the second computing model.
Khare does not appear to explicitly disclose:
“…the container API being defined by a container infrastructure;”
However, Deshpande teaches:
“….the container API being defined by a container infrastructure;” (Paragraph [0038] discloses a container infrastructure that defines its own API and receives provisioning requests through it, “Backend computing environment 104 may further include a container manager 110. Container manager 110 may provide an application programming interface (API) endpoint for host 108. For example, container manager 110 may be accessible to users via a REST (representational state transfer) API (application programming interface) or any other client-server communication protocol…Container manager 110 may receive provisioning requests submitted (e.g., via REST API calls) and provision a container 114 based on the provisioning requests.” Paragraph [0045] further discloses that the infrastructure’s API includes APIs dedicated to resource allocation, “Backend computing environment 104 may further include a set of APIs 204 having one or more workflows 208…For example, workflow 208 may include an allocation phase and APIs 204 may include a given subset of APIs corresponding to the allocation phase.” Paragraph [0040] discloses that a single provisioned container may include more than one selected model, “In some examples, each container image 115 may further include one or more machine learning models.” Paragraph [0010] further discloses that the container infrastructure selects among a first and a second machine learning model of different types held in a single library, “A computing system generates a first machine learning model of a first type. The computing system stores the first machine learning model in the library. The computing system generates a second machine learning model of a second type. The computing system stores the second machine learning model in the library. For each machine learning model, the computing system generates a plugin specific to the respective machine learning model.”)
Therefore, Claim 21 is rejected.
With respect to Claim 22:
Khare-Deshpande teaches:
"The method of claim 21, further comprising: receiving a second model request including one or more second request parameters that are different from the one or more first request parameters, the one or more second request parameters including at least one selected from a group consisting of a type of computing model, a type of computing model, a processing characteristic, and a data characteristic;" (Khare Paragraph [0078] discloses that a query request for a second model is received, “a request is received for a listed repository item (algorithm/model/data/pipeline/notebook) suggestion. For example, a query such as that detailed with respect to FIG.2 is received” Khare Paragraph [0036] further discloses the specific parameters associated with this request, “A query may include one or more of: a category (and subcategory), data information (format of what is available to the requester), resource availability (what resources are available for use such as a processor type), timing information (such as desired latency), an indication of a pipeline that the model/algorithm/data is to be used for), accuracy desired, type of content desired (algorithm, model, and/or data), etc.”) Examiners Note: As per, paragraph [0088] and Fig. 9, the second model request parameters are different from the first, in that the outputted data of the first model, (Y) has been manipulated to reflect changes and variance, thus the request, parameters, and requirements to train/invoke
PNG
media_image1.png
431
1243
media_image1.png
Greyscale
the second computing model are not the same as the first.
“selecting at least a third computing model based upon the second model request;” (Khare Paragraph [0078] discloses determining and selecting a specified third computing model based on the second model request, “At 801, in some embodiments, a request is received for a listed repository item (algorithm/model/data/pipeline/notebook) suggestion. For example, a query such as that detailed with respect to FIG.2 is received.” Khare Paragraph [0079] describes the selecting of a third computing model based on the second model request (query), “Using the details of the query, one or more listed repository items that may meet the desires of the query is determined and provided to the requester at 803.”) Examiner’s Note: As per Fig. 9, the third computing model (model 3 931) [0089] would have already been selected if a data conditioning algorithm/model 929 needed to condition the output parameters/data of model 2 to then be utilized for a third model. A request/selecting of any number of models would have already been predetermined and decided prior to this point in processing. The conditioning of input/output data differentiates the characteristics of the parameters for each individualized computing model.
“compiling a second container request based on the second model request and the third computing model, the second container request including a second request using the container API;” (Khare Paragraph [0070] discloses compiling a second container request, “A request to list a container is received at 701. For example, a producer requests that a container in the model/algorithm container registry 105 or package published in the source control service 107 is to be listed by the publishing listing agent 125 as available for third party use.” Khare Paragraph [0082] mentions that a container request incorporates the selected computing model(s) that were previously determined by the model request(s), “In some embodiments, the selected algorithm or model is a part of a container and a copy of the container is allocated to be executed.” Khare Paragraphs [0032] and [0082] disclose that the request for the selected model is assembled and sent through API gateway 103 to execution service 111, as set forth for claim 1. Deshpande discloses that each provisioning request is submitted “via REST API calls to container manager 110” {[0057] of Deshpande}.) Examiner’s Note: The request that provides the location of model 3 (931) to execution service 111 is a second container request compiled based on the second model request and the third computing model.
“transmitting the second container request to a container infrastructure;” (Khare Paragraph [0068] discloses the transmission of the second container request after it’s been fulfilled, “When the verification was successful, the package is published in the model/algorithm container registry 105. Published containers are available to the publishing/listing agent 125 to be served as a potential result to a code or data query”. Khare Paragraph [0082] discloses pipeline infrastructure (container infrastructure) that takes in all container request related information in order to perform a particular task, “The selected content may be a part of a pipeline and, in those cases, resources are allocated for the pipeline if it is ready for execution (for example, contains only models). For example, a request to perform one or more of these acts is received by the model/algorithm/data API frontend 109 which calls on the publishing/listing agent 125 to provide necessary information to an execution service 111 (such as a location of the selected algorithm, model, data, pipeline, and/or notebook in algorithm/model/data store 123) which then allocates execution resources 113 including compute resources 117 and storage 115.”)
“and causing the container infrastructure to allocate the one or more resources for running the third computing model.” (Khare Paragraph [0082] discloses that, for the selected model, execution service 111 “then allocates execution resources 113 including compute resources 117 and storage 115.” Khare Paragraph [0083] discloses that the allocated resources are selected based on the newly selected model, “These different resources are selected based on the information of the listing (such as suggested resource requirements, latency, etc.). Further, in some embodiments, different resources are selected based on what hardware resources are available to the requesting user.”) Examiner’s Note: Under the broadest reasonable interpretation, “the one or more resources” are the execution resources 113 (compute resources 117 and storage 115) of the container infrastructure, which Khare allocates for running whichever model is selected, including model 3 (931), which occupies its own stage of post-training pipeline 921 ([0089]).
Khare does not appear to explicitly disclose:
“the second container request including a second request using the container API;”
However, Deshpande teaches:
“the second container request including a second request using the container API;” (Paragraph [0038] discloses that the same container API receives multiple provisioning requests, “Container manager 110 may receive provisioning requests submitted (e.g., via REST API calls) and provision a container 114 based on the provisioning requests.” Paragraph [0010] discloses that requests for different types of models are submitted through that API, “A computing system generates a first machine learning model of a first type. … The computing system generates a second machine learning model of a second type. … The computing system receives an application programming interface (API) call to provision a container that includes one of the first machine learning model or the second machine learning model.” Paragraph [0046] further discloses that each request designates the model type to be provisioned, “the type of machine learning model included in container image 115 may be designated by a user or application running on client device in one or more parameters included in the API call from client device 102,” and that “container manager 110 may provide a machine learning model agnostic functionality that allows client device 102 to request a particular type of machine learning model that container manager 110 may leverage to provision a container 114 having the particular machine learning model type pre-loaded therein.” Paragraph [0057] discloses that each such request is submitted through the container API, “client device 102 may submit a provisioning request via REST API calls to container manager 110 for the provisioning of one or more containers 114.” Paragraph [0071] further discloses that additional requests cause the infrastructure to allocate additional resources, “Increasing the number of containers 114 executing on one or more hosts 108 in backend computing environment 104 may increase the number of resources accessible to the machine learning algorithm executing thereon.”) Examiner’s Note: Because Deshpande’s container manager is model agnostic and receives provisioning requests for models of different types through the same REST API, a second request for a different (third) computing model is a second request using the same container API.
Therefore, Claim 22 is rejected.
Claim(s) 5-8, and 17-19 are rejected under 35 U.S.C. 103 as being unpatentable over Khare et. Al, (U.S Patent Application Publication No. US 20220129334 A1, filed on January 10, 2022, hereinafter "Khare"), in view of Deshpande, further in view of Hou et. Al, (U.S Patent Application Publication No. US 20210325861 A1, filed on June 25, 2021, hereinafter "Hou").
With respect to Claims 5 and 17:
Khare-Deshpande does not appear to explicitly disclose:
“The method of claim 1, wherein the container infrastructure is configured to instantiate at least one of the one or more selected computing models.”
However, Hou teaches:
“The method of claim 1, wherein the container infrastructure is configured to instantiate at least one of the one or more selected computing models.” (Paragraph [0085] discloses a process for instantiating a selected computing model, “FIG. 6 is a flowchart representative of example machine readable instructions and/or example operations 600 that may be executed and/or instantiated by processor circuitry to output candidate model combinations based on a model update process.” Paragraph [0088] further reinforces that an instantiation step can be done on a selected model (first artificial intelligence model), “The machine-readable instructions and/or operations 700 of FIG. 7 begin at block 702, at which the example intelligent trigger circuitry 506 receives at least one of four sources of data . For example, the example intelligent trigger circuitry 506 may receive at least one of four sources of data by accessing either a metric baseline, an output from a first artificial intelligence model operating…”)
Khare and Hou are analogous art and in the same field of invention because both references pertain to the utilization of inference-based potential model selection for the purpose of future deployment in dynamically changing computing environments. While Khare teaches gathering a model request to facilitate a container request for a selected model, Hou teaches instantiating said selected model with respect to the suggested requirements. 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 implement the base reference of Khare (optimizing repositories to refine the selection of necessary machine learning models), as modified by Deshpande, with the teachings of Hou (updating/generating feedback-based instantiated artificial intelligence models for autonomous environments) in order to eliminate configuration discrepancies, improve version rollback/control tracking, and standardize overall inference latency. One of ordinary skill in the art would be motivated to do so because by integrating Hou’s framework into the methods of Khare one would be able to “provide the ability to serve and respond to multiple applications of the use cases 205 (e.g., object tracking, video surveillance, connected cars, etc.) in real-time or near real time, and meet ultra-low latency requirements for these multiple applications. These advantages enable a whole new a class of applications (Virtual Network Functions (VNFs), Function as a Service (FaaS), Edge as a Service (EaaS), standard processes, etc.)” {[0043] of Hou}.
Therefore Claims 5 and 17 are rejected.
With respect to Claims 6 and 18:
Khare-Deshpande teaches:
“The method of claim 5, wherein the container request includes one or more configurations or one or more connection requirements of the one or more selected computing models; wherein the container infrastructure is configured to…” (Khare Paragraph [0080] references the selecting of a computing model, “At some point a request for a selected repository item is received at 805…When the requester already knows which algorithm/model/data is desired,” Khare Paragraph [0081] further discloses that container requests to transmit models to the container infrastructure (ML pipeline) can embody connection related protocols, “In some embodiments, a request to add the repository item to a ML pipeline is received at 807. For example, when the publishing/listing agent 125 is helping a user build a pipeline to perform a task (or tasks) this type of request occurs. The publishing/listing agent 125 will evaluate the pipeline as it exists and make the necessary connections within the pipeline at 809.” Khare Paragraph [0024] further teaches that the entity (the publishing/listing agent 125) that receives and responds to the container request also manages certain configuration requirements associated with the selected computing model as per paragraph [0065], “In some embodiments, requirements for using the code or data are received from the producer at 603. For example, the publishing/listing agent 125 is provided with hardware configurations to use, or encryption to adhere to, etc.”)
Khare-Deshpande does not appear to explicitly disclose:
“…instantiate the at least one of the one or more selected computing models based at least in part upon the one or more configurations or the one or more connection requirements.”
However, Hou teaches:
“…instantiate the at least one of the one or more selected computing models based at least in part upon the one or more configurations or the one or more connection requirements.” (Paragraph [0085] references a figure that showcases the process of instantiating a selected computing model, “FIG.6 is a flowchart representative of example machine readable instructions and/or example operations 600 that may be executed and/or instantiated by processor circuitry to output candidate model combinations based on a model update process.” Paragraph [0155] further teaches the instantiation based on configuration and connection requirements, “the FPGA circuitry 1300 of the example of FIG.13 includes interconnections and logic circuitry that may be configured and/or interconnected in different ways after fabrication to instantiate, for example, some or all of the machine-readable instructions represented by the flowchart of FIGS. 6-7.”)
Therefore, Claims 6 and 18 are rejected.
With respect to Claim 7:
Khare-Deshpande does not appear to explicitly disclose:
“The method of claim 5, wherein the container request includes metadata corresponding to the one or more selected computing models; wherein the container infrastructure is configured to instantiate the at least one of the one or more selected computing models based at least in part upon the metadata.”
However, Hou teaches:
“The method of claim 5, wherein the container request includes metadata corresponding to the one or more selected computing models; wherein the container infrastructure is configured to instantiate the at least one of the one or more selected computing models based at least in part upon the metadata.” (Paragraph [0056] discloses metadata associated with a particular selected computing model, “The example model update controller circuitry 410 includes access to a database 412 of sensor data or environmental metadata. The other example model update controller circuitry 418 includes access to a database 416 of sensor data or environmental metadata. The other example model update controller circuitry 418 produces an ensemble (e.g., at least one) of artificial intelligence models in the database 420 which is communicated to the example network 406 and distributed to the example model update controller circuitry 410. Paragraph [0088] teaches instantiation of the selected model based on metadata, “FIG . 7 is a flowchart representative of example machine readable instructions and/or example operations 700 that may be executed and/or instantiated by processor circuitry to output candidate model combinations based on a model update process. The machine-readable instructions and/or operations 700 of FIG. 7 begin at block 702, at which the example intelligent trigger circuitry 506 receives at least one of four sources of data. For example, the example intelligent trigger circuitry 506 may receive at least one of four sources of data by accessing either a metric baseline, an output from a first artificial intelligence model operating on the first factory production line, metadata…”)
Therefore, Claims 7 is rejected.
With respect to Claims 8 and 19:
Khare-Deshpande does not appear to explicitly disclose:
“The method of claim 1, wherein the container infrastructure is configured to update at least one of the one or more selected computing models.”
However, Hou teaches:
“The method of claim 1, wherein the container infrastructure is configured to update at least one of the one or more selected computing models.” (Paragraph [0032] discloses selecting (deploying) a computing model to be updated, “The model is stored at the model repository as described in FIG. 5. The model may then be executed by the intelligent deployment circuitry as described in FIG. 5. In some examples, multiple models are deployed and evaluated by the intelligent deployment circuitry as described in FIG. 5.” Paragraph [0034] further discloses updating the selected computing model, “If the feedback indicates that the accuracy of the deployed model is less than a threshold or other criterion, training of an updated model can be triggered using the feedback and an updated training data set, hyperparameters, etc., to generate an updated, deployed model.”)
Therefore, Claims 8 and 19 are rejected.
Claims 11 and 12 are rejected under 35 U.S.C. 103 as being unpatentable over Khare et. Al, (U.S Patent Application Publication No. US 20220129334 A1, filed on January 10, 2022, hereinafter "Khare"), in view of Deshpande, further in view of Rogers et. Al, (U.S Patent Application Publication No. US 20210117859 A1, filed on September 9, 2020, hereinafter "Rogers").
With respect to Claim 11:
Khare-Deshpande teaches:
“The method of claim 10, wherein the processing pipeline includes a third computing model operating in parallel with respect to the first computing model.” (Khare Paragraph [0086] references a processing pipeline that includes a first model “The pre-training pipeline 901 includes a first model a (model 1) 903…” Khare Paragraph [0088] further discloses that a third computing model is present on the same pipeline and relevant to the first computing model which has undergone training by algorithm 2, “The output of the model generated by training algorithm 2 905 is in the format of Z. As shown, the input into selected model 3 911 is Z'. In this example, the pre-training pipeline 901 requires an intermediary, data conditioning algorithm/model 909, between what will be model 2 and model 3 911, which conditions Z to be Z'. Z' is then fed into model 3 911.”)
Khare-Deshpande does not appear to explicitly disclose:
“…computing model operating in parallel...”
However, Rogers teaches:
“…computing model operating in parallel...” (Paragraph [0112] discloses that artificial intelligence related processes/services can run in parallel with each other, “to aid in processing or execution of applications or containers in pipelines, services 1320 may be leveraged. In at least one embodiment, services 1320 may include compute services, artificial intelligence (AI) services…functionality provided by services 1320 may run dynamically and more efficiently, while also scaling well by allowing applications to process data in parallel (e.g., using a parallel computing platform 1430 (FIG. 14).” Paragraph [0128] mentions that AI services can encompass multiple models in relation to execution of said parallel services, “, AI services 1418 may leverage AI system 1424 to execute machine learning model (s) (e.g., neural networks, such as CNNs) for segmentation, reconstruction, object detection, feature detection, classification, and/or other inferencing tasks. In at least one embodiment, applications of deployment pipeline (s) 1410 may use one or more of output models 1316…”)
Khare and Rogers are analogous art and in the same field of invention because both references pertain to the utilization of inference-based potential model selection for the purpose of future deployment in robust operational architectures. While Khare teaches generating a processing pipeline for efficiency in machine learning model task performance, Rogers teaches fostering the means for machine learning models to co-conduct operations in parallel with one another. 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 implement the base reference of Khare (improving machine learning pipeline implementation through the use of model repositories), as modified by Deshpande (automating and simplifying the deployment of machine learning models using container technology), with the teachings of Rogers (improving machine learning model integrity to run seamlessly) in order to increase prediction accuracy whilst reducing training time and heightening data as well as model parallelism. One of ordinary skill in the art would be motivated to do so because by integrating Rogers’ framework into the methods of Khare one would be able to “provide an ability to update a model with zero downtime, or near-zero downtime with no data loss, and without the need to update or restart the application” {[0023] of Rogers}.
Therefore, Claim 11 is rejected.
With respect to Claim 12:
Khare-Deshpande teaches:
“The method of claim 9, wherein the processing pipeline includes at least a first computing model and a second computing model operating in parallel.” (Khare Paragraph [0086] references a processing pipeline that includes a first model “The pre-training pipeline 901 includes a first model a (model 1) 903…” Khare Paragraph [0089] further discloses that a second computing model is present on the same pipeline and operating parallel to the first computing model “The output (Y) of the model 1 903 is an input into model 2 925.”
Khare-Deshpande does not appear to explicitly disclose:
“…computing model operating in parallel...”
However, Rogers teaches:
“…computing model operating in parallel...” (Paragraph [0112] discloses that artificial intelligence related processes/services can run in parallel with each other, “to aid in processing or execution of applications or containers in pipelines, services 1320 may be leveraged. In at least one embodiment, services 1320 may include compute services, artificial intelligence (AI) services…functionality provided by services 1320 may run dynamically and more efficiently, while also scaling well by allowing applications to process data in parallel (e.g., using a parallel computing platform 1430 (FIG. 14).” Paragraph [0128] mentions that AI services can encompass multiple models in relation to execution of said parallel services, “, AI services 1418 may leverage AI system 1424 to execute machine learning model (s) (e.g., neural networks, such as CNNs) for segmentation, reconstruction, object detection, feature detection, classification, and/or other inferencing tasks. In at least one embodiment, applications of deployment pipeline (s) 1410 may use one or more of output models 1316…”)
Therefore, Claim 12 is rejected.
Claims 13, 20, 23, and 24 are rejected under 35 U.S.C. 103 as being unpatentable over Khare et. Al, (U.S Patent Application Publication No. US 20220129334 A1, filed on January 10, 2022, hereinafter "Khare"), in view of Deshpande, further in view of Lee et. Al, (U.S Patent Application Publication No. US 20230315856 Al, filed on March 31, 2022, hereinafter "Lee").
With respect to Claims 13 and 20:
Khare-Deshpande teaches:
“The method[system] of claim 1[14], wherein the one or more selected computing models…” (Khare Paragraph [0078] discloses determining and selecting a specified computing model based on the model request, “At 801, in some embodiments, a request is received for a listed repository item (algorithm/model/data/pipeline/notebook) suggestion. For example, a query such as that detailed with respect to FIG.2 is received.” Khare Paragraph [0079] describes the selecting of a computing model based on the model request (query), “Using the details of the query, one or more listed repository items that may meet the desires of the query is determined and provided to the requester at 803.”)
Khare-Deshpande does not appear to explicitly disclose:
“…include a large language model.”
However, Lee teaches:
“…include a large language model.” (Paragraph [0066] teaches that the machine learning/computing models can be large language models, “Each of the ML models 212 can be a language model trained to receive natural language data and interpret the natural language data based on user needs and context. The ML models 212 can be implemented using any suitable method, process, or framework.”)
Khare and Lee are analogous art and in the same field of invention because both references pertain to the utilization of inference-based potential model selection for the purpose of future deployment in dynamically changing computing environments. While Khare teaches selecting a particular artificial intelligence model for refinement and implementation based on relative requirement parameters, Lee teaches ensuring the artificial intelligence models embody large language model principles and capabilities. 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 implement the base reference of Khare (employing a repository of diverse and varied machine learning models), as modified by Deshpande (generate and provision a customized container (or virtual machine) that has a specific machine learning model pre-loaded based on parameters provided by an end user), with the teachings of Lee (augmenting complex natural language-based training/querying) in order to detect unstructured data nuances, provide flexible architecture versatility, and increase time-efficient automation as well as deep-reasoning abilities. One of ordinary skill in the art would be motivated to do so because by integrating Lee’s framework into the methods of Khare one could deliver a beneficial outcome, “resulting in an improved, intuitive, user interface where users can use simple natural language to interact and manipulate complex management systems instead of having to go through rigorous time-consuming training of complicated technical jargon or learning to navigate intricate menu trees with functions listed only to find the correct function to perform each task intended” {[0077] of Lee}.
Therefore, Claims 13 and 20 are rejected.
With respect to Claim 23:
Khare-Deshpande does not appear to explicitly disclose:
“The method of claim 21, wherein the first computing model is a first large language model.”
However, Lee teaches:
“The method of claim 21, wherein the first computing model is a first large language model.” (Paragraph [0103] teaches that the first computing model can be a large language model, “The first ML model can be a Generative Pre-trained Transformer 3 (GPT-3) model, or a GPT-2 model, OpenAI Davinci Codex, OpenAI Cushman codex, GPT-NE02.7B and finetuned GPT-NE02.7B, or any other suitable language model trained on corpuses of language data.”)
Therefore, Claim 23 is rejected.
With respect to Claim 24:
Khare-Deshpande does not appear to explicitly disclose:
“The method of claim 23, wherein the second computing model is a second large language model.”
However, Lee teaches:
“The method of claim 23, wherein the first computing model is a second large language model.” (Paragraph [0105] teaches that the second computing model can be a large language model, “The second ML model can be configured to generate a set of natural language phrases semantically related to the reference input. The second ML model can be a GPT-3 model, GPT-2 model, OpenAI Davinci Codex, OpenAI Cushman codex, GPT-NE02.7B and finetuned GPT-NE02.7B, or any other suitable language model trained on large corpuses of language data.”)
Therefore, Claim 24 is rejected.
Response to Amendment
Claim Rejections - 35 USC § 101
Applicant’s amendments and arguments July 21st, 2026, with respect to the rejection of claims 1-24 under 35 U.S.C. § 101 have been fully considered. The rejection of claims 1, 3-14, and 16-24 under 35 U.S.C. § 101 is withdrawn. The rejection of claims 2 and 15 is moot in view of their cancellation.
Although claim 1 still recites selecting one or more computing models based upon the model request, which is a mental process (MPEP § 2106.04(a)(2)(III)), the amended independent claims now positively require compiling a container request using a container API defined by a container infrastructure, transmitting that request to the infrastructure, and causing the infrastructure to allocate one or more resources for running the selected computing models. Allocation was previously recited only in dependent claims 2 and 15 as a capability of the infrastructure. Considered as a whole, the amended claims apply the selection to provision and run the selected models on the container infrastructure, which integrates the judicial exception into a practical application. MPEP § 2106.04(d). The examiner does not rely on the benefits described in paragraph [0018] of the specification, which are not reflected in the claims. Independent claims 14 and 21 recite the same limitations, and the dependent claims are eligible for the same reasons.
Claim Rejections - 35 USC § 102
Applicant’s amendments and arguments filed July 21st, 2026, with respect to the rejection of claims 1-24 under 35 U.S.C. §§ 102 and 103 have been fully considered but are not persuasive. The rejection of claims 2 and 15 is moot in view of their cancellation.
Claims 1, 3-14, and 16-24 are rejected under 35 U.S.C. § 103 over Khare in view of Deshpande (US20190279114A1), alone or further in view of Hou, Rogers, or Lee, in a new ground of rejection necessitated by Applicant’s amendment, further rendering said arguments moot.
Claim Rejections - 35 USC § 103
A: Applicant argues that the “request to list a container” at step 701 of Khare ([0070]) is part of the separate FIG. 7 listing process, which takes place before the FIG. 8 query process. Applicant concludes that the request to list a container is not compiled based on the query or on the listed repository items that meet the query.
The examiner does not agree. Applicant’s argument treats the request to list a container at step 701 [0070] as if it alone were relied upon as the claimed container request. It is not. The rejection relies on [0070] together with [0082] (and for claims 21 and 22 [0032]) and the container request reads on the combined disclosure.
First, step 701 establishes the container and its location for later use: “typically, the request 701 includes this location” of the container in container registry 105 [0071]. The listed container is then made available for query results, as “a query for the code will return the API wrapped microservice which can be executed on execution resources 113” [0075]. FIG. 8 expressly begins with this publishing and listing at step 800 [0077]. The FIG. 7 and FIG. 8 processes are therefore successive stages of a single flow, not unrelated methods. Applicant concedes as much in stating that “the requester cannot request for a container…unless the container is already listed.”
Second, after the query is received 801, [0078] and the model is selected based on it 803, [0079], the request to execute the selected model is compiled. API frontend 109 “calls on the publishing/listing agent 125 to provide necessary information to an execution service 111 (such as a location of the selected algorithm, model, data, pipeline, and/or notebook in algorithm/model/data store 123),” and “the selected algorithm or model is a part of a container, and a copy of the container is allocated to be executed” [0082]. That request is based on the model request, because it issues only in response to the query and selection. It is based on the selected computing model, because it carries the selected model’s location, the same container location established at step 701.
Claim 1 requires that the container request be compiled “based on the model request and the one or more selected computing models.” It does not require that every step associated with the container, including making the container available, occur after the model request is received. Accordingly, Khare’s [0070] and [0082], read together, teach compiling a container request based on the model request and the one or more selected computing models.
B: Applicant argues that Khare’s API frontend 109 is separate from container registry 105 and is not a container API defined by a container infrastructure.
This argument is moot. The rejection does not rely on API frontend 109. Khare is relied upon for API gateway 103, which “takes in a request at circle 8 to start execution (e.g., inference) or training by an execution service 111” [0032], and for the containerized model “wrapped with an API” [0074]. Deshpande is relied upon for the API being defined by the container infrastructure (Deshpande [0038], [0045]), in a new ground of rejection necessitated by amendment.
C: Applicant argues that the amended claims distinguish over Khare by reciting “causing the container infrastructure to allocate one or more resources for running the one or more selected computing models.”
The examiner does not agree. This limitation was incorporated from cancelled claims 2 and 15, which were rejected over Khare and not separately argued. Khare’s execution service 111 “allocates execution resources 113 including compute resources 117 and storage 115” to execute the selected model [0082].
D: Applicant argues that Hou, Rogers, and Lee do not disclose a container API defined by a container infrastructure.
The examiner does not agree. Those references are not relied upon for that limitation, which is taught by Khare in view of Deshpande. One cannot show non-obviousness by attacking references individually where the rejections are based on combinations of references. In re Keller, 642 F.2d 413, 426 (CCPA 1981).
Conclusion
THIS ACTION IS MADE FINAL. 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 NOOR F CHEEMA whose telephone number is (571)272-9642. The examiner can normally be reached Monday-Friday 7:30am-5:00pm alternative Fridays off.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Mariela Reyes can be reached at (571) 270-1006. 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.
/N.F.C./ Examiner, Art Unit 2142
/HAIMEI JIANG/ Primary Examiner, Art Unit 2142
