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 .
Claim Rejections - 35 USC § 102
The following is a quotation of the appropriate paragraphs of 35 U.S.C. 102 that form the basis for the rejections under this section made in this Office action:
A person shall be entitled to a patent unless –
(a)(1) the claimed invention was patented, described in a printed publication, or in public use, on sale, or otherwise available to the public before the effective filing date of the claimed invention.
Claim(s) 1-5, 7, 9 and 11 is/are rejected under 35 U.S.C. 102(a)(1) as being anticipated by Le U.S. PAP 2022/0092269 A1.
Regarding claim 1 Le teaches a computer-implemented method for executing system actions during a conversation between a human user and an autonomous conversational system (Methods and systems are described herein for generating dynamic conversational responses, see abstract), the method comprising:
receiving a user message from the human user via a communication channel (receives a first user action during a conversational interaction with a user interface, see par. [0003]);
providing the user message as input to a first generative language model and receiving as output an indication of a user intent (the system to determine both quickly (i.e., in real-time or near real-time) and accurately the intent, goal, or motivation of a user when the user interacts with the system, see par. [0003]; he BERT model includes pre-training contextual representations including Semi-supervised Sequence Learning, Generative Pre-Training, see par. [0048]);
determining, by a dialog management model based on the user intent and conversation context, that a system action is required ( the system must aggregate information about the user action, information about the user, and/or other circumstances related to the user action (e.g., time of day, previous user actions, current account settings, etc.) in order to determine a likely intent of the user, see par. [0003]);
executing the system action to obtain response data ( the method or system may generate dynamic conversational responses using multiple machine learning models,, see par. [0008]), wherein executing comprises:
retrieving required parameters from the conversation context, performing the system action using the required parameters, and storing results of the system action in conversation context variables (he system may determine, based on the first user action, a first feature input for a first machine learning model, wherein the first machine learning model comprises a plurality of attention layers functioning in parallel. The system may input the first feature input into the first machine learning model to generate a first output from the first machine learning model, wherein the first machine learning model comprises a plurality of attention layers functioning in parallel. The system may input the first output into a second machine learning model to generate a second output, see par. [0008]; the system may generate data points such as demographic segments (age, gender, profession, household income), temporal aspects (time of day, season, events), geolocation, and other behavioral data during a device session into order to determine insights into the specifics of the context of usage of a particular digital product or service, see par. [0025]);
generating, by a second generative language model using the response data, a response message to the human user (The system may then select a dynamic conversational response from a plurality of dynamic conversational responses based on the second output, see par. [0008]);
and transmitting the response message to the human user via the communication channel (The system may generate, at the user interface, the dynamic conversational response during the conversational interaction, see par. [0008]).
Regarding claim 2 Le teaches the method of claim 1, wherein executing the system action comprises:
retrieving a database query configuration from stored system actions ( inputting a query into user interface 100, see par. [0024]);
executing the database query on a connected database system using the required parameters to obtain the response data (retrieving data about the user, retrieving data from other sources, monitoring user actions, and/or other steps in order to generate a feature input, see par. [0024]);
validating the response data received from the database system (The system may then retrieve time-dependent user account information for the first user during the device interface session with the one or more user interfaces, see par. [0026]);
and storing the validated response data in the conversation context variables (the feature input may include a vector that describes various information about a user, a user action, and/or a current or previous interaction with the user. The system may further select the information for inclusion in the feature input based on a predictive value, see par. [0026]).
Regarding claim 3 Le teaches the method of claim 1, wherein executing the system action comprises:
retrieving an API call configuration from stored system actions (include conversation details such as information about a current session, including a channel or platform, e.g. desktop web, iOS, mobile, a launch page (e.g., the webpage that the application was launched from), a time of launch, or activities in a current or previous session before launching the application , see par. [0027]);
executing the API call to a third-party system using the required parameters to obtain the response data (the types of information (e.g., user actions and/or time-dependent information) may include insights about users, provided to the application (e.g., via an API) from one or more sources such as a qualitative or quantitative representations (e.g., a percent) of a given activity, see par. [0029]);
validating the response data received from the third-party system (The system may obtain this data in real-time for model prediction through enterprise APIs, see par. [0028]);
and storing the validated response data in the conversation context variables ( the system may process transaction data. For example, the record data may include a paper or electronic record containing information about the transaction, see par. [0031]).
Regarding claim 4 Le teaches the method of claim 1, wherein the conversation context comprises:
previously exchanged messages between the human user and the autonomous conversational system (he system must aggregate information about the user action, information about the user, and/or other circumstances related to the user action (e.g., time of day, previous user actions, current account settings, etc., see par. [0003]);
results from previously executed system actions (results, see par. [0041]);
extracted parameter values from the user message (multiple datasets, such as a training dataset and a test dataset. Each of the plurality of datasets (e.g., inputs 204) may include data subsets related to user data, contact strategies, and results, see par. [0041]));
and conversation state information for use by one or more of a natural language understanding model, dialog management model, and generative language model (a first type of data (e.g., a user action) may include conversation details such as information about a current session, see par. [0027]; The system may then translate the first type of data into data arrays of numbers using natural language processing, see par. [0023]).
Regarding claim 5 Le teaches the method of claim 1, wherein generating the response message comprises:
selecting a response template based on the system action results ( select a subset of dynamic conversational responses from a plurality of dynamic conversational responses, see par. [0007]);
filling placeholder parameters in the template with values from the response data (system may generate dynamic conversational responses using multiple machine learning models, the method comprising. For example, the system may receive a first user action during a conversational interaction with a user interface. The system may determine, based on the first user action, a first feature input for a first machine learning model, see par. [0008]);
and generating a natural language response with the filled template using the second generative language model (The system may input the first output into a second machine learning model to generate a second output, wherein the second machine learning model comprises a plurality of convolutional neural networks and a Leaky Rectified Linear Unit (“LeakyReLU”) activation function. The system may then select a dynamic conversational response from a plurality of dynamic conversational responses based on the second output, see par. [0008]).
Regarding claim 7 Le teaches the method of claim 1, wherein the system action comprises:
retrieving user account information (data such as time-dependent user account information, see par. [0023]), and wherein executing the system action comprises:
accessing a database containing user account records (The system may then retrieve time-dependent user account information for the first user during the device interface session with the one or more user interfaces, see par. [0026]);
querying the database using user identification parameters extracted from the conversation context ( the feature input may include a vector that describes various information about a user, a user action, and/or a current or previous interaction with the user. The system may further select the information for inclusion in the feature input based on a predictive value. The information may be collected actively or passively by the system and compiled into a user profile, see par. [0026]);
and retrieving account details required for generating the response message (a second type of data (e.g., a time-dependent information) may include user account information, such as types of accounts the user has, other accounts on file, such as bank accounts for payment, information associated with accounts, such as credit limit, current balance, due date, recent payments, recent transactions. The system may obtain this data in real-time for model prediction through enterprise APIs., see par. [0028]).
Regarding claim 9 Le teaches the method of claim 1, wherein determining that a system action is required comprises:
analyzing the user intent and conversation context to identify required system actions (The system may comprise a model that predicts an intent of a user. For example, the system may determine if a customer intends to make a credit card payment. To do so, the system may monitor a first type of data (e.g., user actions in interface 100) and/or other types of data such as time-dependent user account information, see par. [0023]);
retrieving a stored action configuration corresponding to the identified required system actions (he first type may include a set of text pages, reflecting the contents of the internet drive menu pages. The second type may include a set of numerical and categorical values. The system may then translate the first type of data into data arrays of numbers using natural language processing., see par. [0023]);
validating that all required parameters are available in the conversation context (retrieving data about the user, retrieving data from other sources, monitoring user actions, see par. [0024]);
and selecting between multiple available system actions based on conversation state and parameter availability (the system may take one or more steps to generate dynamic conversational responses, and/or select a subset of dynamic conversational responses, see par. [0024]).
Regarding claim 11 Le teaches the method of claim 1, wherein the first generative language model and the second generative language model are either:
(i) a single generative language model trained to perform both user intent determination and response message generation tasks,
or (ii) two separate generative language models, wherein: the first generative language model is specifically trained to determine user intent from received user messages, and the second generative language model is specifically trained to generate response messages using response data from executed system actions (The system may input the first feature input into the first machine learning model to generate a first output from the first machine learning model, wherein the first machine learning model comprises a plurality of attention layers functioning in parallel. The system may input the first output into a second machine learning model to generate a second output, wherein the second machine learning model comprises a plurality of convolutional neural networks and a Leaky Rectified Linear Unit (“LeakyReLU”) activation function. , see par. [0008]).
Regarding claim 12 Le teaches a system for executing system actions during a conversation between a human user and an autonomous conversational system (Methods and systems are described herein for generating dynamic conversational responses, see abstract), the system comprising:
a communication interface configured to receive user messages from the human user ( dynamic interface options may facilitate an interactive exchange with users. The interactive exchange may include the system responding to one or more user actions (or inactions) and/or predicting responses prior to receiving a user action, see par. [0003]);
one or more processors (see par. [0035]);
and memory storing instructions that, when executed by the one or more processors (The control circuitry may comprise any suitable processing, storage, and/or input/output circuitry, see par. [0035]), cause the system to:
receive a user message from the human user via a communication channel (receives a first user action during a conversational interaction with a user interface, see par. [0003]);
provide the user message as input to a first generative language model and receiving as output an indication of a user intent (the system to determine both quickly (i.e., in real-time or near real-time) and accurately the intent, goal, or motivation of a user when the user interacts with the system, see par. [0003]; he BERT model includes pre-training contextual representations including Semi-supervised Sequence Learning, Generative Pre-Training, see par. [0048]);
determine, by a dialog management model based on the user intent and conversation context, that a system action is required ( the system must aggregate information about the user action, information about the user, and/or other circumstances related to the user action (e.g., time of day, previous user actions, current account settings, etc.) in order to determine a likely intent of the user, see par. [0003]);
executing the system action to obtain response data ( the method or system may generate dynamic conversational responses using multiple machine learning models,, see par. [0008]), wherein executing comprises:
retrieve required parameters from the conversation context, performing the system action using the required parameters, and storing results of the system action in conversation context variables (he system may determine, based on the first user action, a first feature input for a first machine learning model, wherein the first machine learning model comprises a plurality of attention layers functioning in parallel. The system may input the first feature input into the first machine learning model to generate a first output from the first machine learning model, wherein the first machine learning model comprises a plurality of attention layers functioning in parallel. The system may input the first output into a second machine learning model to generate a second output, see par. [0008]; the system may generate data points such as demographic segments (age, gender, profession, household income), temporal aspects (time of day, season, events), geolocation, and other behavioral data during a device session into order to determine insights into the specifics of the context of usage of a particular digital product or service, see par. [0025]);
generate, by a second generative language model using the response data, a response message to the human user (The system may then select a dynamic conversational response from a plurality of dynamic conversational responses based on the second output, see par. [0008]);
and transmit the response message to the human user via the communication channel (The system may generate, at the user interface, the dynamic conversational response during the conversational interaction, see par. [0008]).
Regarding claim 13 Le teaches the system of claim 12, wherein executing the system action comprises:
retrieving a database query configuration from stored system actions ( inputting a query into user interface 100, see par. [0024]);
executing the database query on a connected database system using the required parameters to obtain the response data (retrieving data about the user, retrieving data from other sources, monitoring user actions, and/or other steps in order to generate a feature input, see par. [0024]);
validating the response data received from the database system (The system may then retrieve time-dependent user account information for the first user during the device interface session with the one or more user interfaces, see par. [0026]);
and storing the validated response data in the conversation context variables (the feature input may include a vector that describes various information about a user, a user action, and/or a current or previous interaction with the user. The system may further select the information for inclusion in the feature input based on a predictive value, see par. [0026]).
Regarding claim 15 Le teaches the system of claim 12, wherein the conversation context comprises: previously exchanged messages between the human user and the autonomous conversational system (he system must aggregate information about the user action, information about the user, and/or other circumstances related to the user action (e.g., time of day, previous user actions, current account settings, etc., see par. [0003]);
results from previously executed system actions (results, see par. [0041]);
extracted parameter values from the user message (multiple datasets, such as a training dataset and a test dataset. Each of the plurality of datasets (e.g., inputs 204) may include data subsets related to user data, contact strategies, and results, see par. [0041]));
and conversation state information for use by one or more of a natural language understanding model, dialog management model, and generative language model (a first type of data (e.g., a user action) may include conversation details such as information about a current session, see par. [0027]; The system may then translate the first type of data into data arrays of numbers using natural language processing, see par. [0023]).
Regarding claim 16 Le teaches the system of claim 12, wherein generating the response message comprises: selecting a response template based on the system action results ( select a subset of dynamic conversational responses from a plurality of dynamic conversational responses, see par. [0007]);
filling placeholder parameters in the template with values from the response data (system may generate dynamic conversational responses using multiple machine learning models, the method comprising. For example, the system may receive a first user action during a conversational interaction with a user interface. The system may determine, based on the first user action, a first feature input for a first machine learning model, see par. [0008]);
and generating a natural language response with the filled template using the second generative language model (The system may input the first output into a second machine learning model to generate a second output, wherein the second machine learning model comprises a plurality of convolutional neural networks and a Leaky Rectified Linear Unit (“LeakyReLU”) activation function. The system may then select a dynamic conversational response from a plurality of dynamic conversational responses based on the second output, see par. [0008]).
Regarding claim 18 Le teaches the system of claim 12, wherein the system action comprises retrieving user account information (data such as time-dependent user account information, see par. [0023]), and wherein executing the system action comprises:
accessing a database containing user account records (The system may then retrieve time-dependent user account information for the first user during the device interface session with the one or more user interfaces, see par. [0026]);
querying the database using user identification parameters extracted from the conversation context ( the feature input may include a vector that describes various information about a user, a user action, and/or a current or previous interaction with the user. The system may further select the information for inclusion in the feature input based on a predictive value. The information may be collected actively or passively by the system and compiled into a user profile, see par. [0026]);
and retrieving account details required for generating the response message (a second type of data (e.g., a time-dependent information) may include user account information, such as types of accounts the user has, other accounts on file, such as bank accounts for payment, information associated with accounts, such as credit limit, current balance, due date, recent payments, recent transactions. The system may obtain this data in real-time for model prediction through enterprise APIs., see par. [0028]).
Regarding claim 19 Le teaches the system of claim 12, wherein determining that a system action is required comprises: analyzing the user intent and conversation context to identify required system actions (The system may comprise a model that predicts an intent of a user. For example, the system may determine if a customer intends to make a credit card payment. To do so, the system may monitor a first type of data (e.g., user actions in interface 100) and/or other types of data such as time-dependent user account information, see par. [0023]);
retrieving a stored action configuration corresponding to the identified required system actions (he first type may include a set of text pages, reflecting the contents of the internet drive menu pages. The second type may include a set of numerical and categorical values. The system may then translate the first type of data into data arrays of numbers using natural language processing., see par. [0023]);
validating that all required parameters are available in the conversation context (retrieving data about the user, retrieving data from other sources, monitoring user actions, see par. [0024]);
and selecting between multiple available system actions based on conversation state and parameter availability (the system may take one or more steps to generate dynamic conversational responses, and/or select a subset of dynamic conversational responses, see par. [0024]).
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.
Claim(s) 6, 8, 10 is/are rejected under 35 U.S.C. 103 as being unpatentable over Le U.S. PAP 2022/0092269 A1, in view of Mazza U.S. PAP 2019/0182382 A1.
Regarding claim 6 Le teaches the method of claim 1, further comprising:
detecting when execution of the system action fails (machine learning model 202 may update its configurations (e.g., weights, biases, or other parameters) based on the assessment of its prediction (e.g., outputs 206) and reference feedback information (e.g., user indication of accuracy, see par. [0042]);
generating an alternative response without using the system action (Updates to the connection weights may, for example, be reflective of the magnitude of error propagated backward after a forward pass has been completed. In this way, for example, the machine learning model 402 may be trained to generate better predictions., see par. [0042]).
However Le does not teach storing information about the failed system action attempt;
and initiating a handoff to a human agent when system action failures exceed a threshold.
IN the same field of endeavor Mazza teaches The interactions may include a plurality of excluded interactions, each of the excluded interactions failing to satisfy the threshold semantic distance, see par. [0019]. Also The contact center system may further include a chatbot service server 170 configured to provide chatbot services to the contact center. FIG. 2 is a block diagram illustrating a chatbot service 170 according to one embodiment of the present invention. The chatbot service server may include a data extraction module 172 configured to extract information from transcripts of prior chat interactions (e.g., stored in the mass storage device 126 and/or the call recording server 158). The extracted data may be provided to a chatbot generation module 174, which is configured to generate a chatbot from the extracted data. The chatbot service server 170 may provide a user interface 176 for human supervisors or designers to control the configuration of the chatbots generated by the chatbot generation module 174, such as by displaying one or more suggestions extracted by the data extraction module 172, allowing the human designer to approve, reject, or edit the suggestions, and to control the final generation of the chatbot, see par. [0074].
It would have been obvious to one of ordinary skill in the art to combine the Le invention with the teachings of Mazza for the benefit of allowing the human designer to approve, reject, or edit the suggestions, and to control the final generation of the chatbot, see par. [0074].
Regarding claim 8 Mazza teaches the method of claim 1, wherein the system action comprises processing a service request, and wherein executing the system action comprises:
calling a third-party API endpoint with required service parameters (The contact center may be an in-house facility to a business or enterprise for serving the enterprise in performing the functions of sales and service relative to the products and services available through the enterprise. In another aspect, the contact center may be operated by a third-party service provider, see par. [0055]);
receiving confirmation of the service request processing (when a customer indicates that he or she would like to reset their password, the agent may respond with the newly reset password, or the agent may confirm the customer's username first (without being asked), see par. [0145]);
and storing the confirmation details in the conversation context variables ( in response to a customer request to reset a password, the chatbot may be designed to always (e.g., deterministically) directly provide the newly reset password, see par. [0146]).
It would have been obvious to one of ordinary skill in the art to combine the Le invention with the teachings of Mazza for the benefit of allowing the human designer to approve, reject, or edit the suggestions, and to control the final generation of the chatbot, see par. [0074].
Regarding claim 10 Le teaches the method of claim 1, wherein the conversation context variables comprise: a conversation identifier uniquely identifying the conversation (conversation details, see par. [0027]);
However Le does not teach turn numbers tracking a sequence of message exchanges; system action results organized by turn number; user intent information from the first generative language model; and parameter values extracted from user messages and system action results.
IN the same field of endeavor Mazza teaches The dialog graph contains many possible flows, mainly due to variability between the agents: the agent may or may not ask for the customer's inquiry reason; based on that reason, he may or may not choose to offer a phone call; and finally, he decides whether to up-sell another voice service product. The pruned dialogue tree organizes the agent's prompts in a consistent order, see par. [0148]. In some embodiments, the intent classification the sentences may be clustered using supervised intent classification using training examples generated by a chatbot designer (e.g., manually or selected from a list of examples), see par. [0128]. he clustering results are projected on the original chats: each sentence is replaced by its cluster's representative (e.g., the most frequently-appearing sentence in the cluster). The sentences occurring in at least a threshold percentage of the interactions are set as the anchors of the topic., see par. [0129].
It would have been obvious to one of ordinary skill in the art to combine the Le invention with the teachings of Mazza for the benefit of allowing the human designer to approve, reject, or edit the suggestions, and to control the final generation of the chatbot, see par. [0074].
Regarding claim 14 Mazza teaches the system of claim 12, wherein executing the system action comprises:
calling a third-party API endpoint with required service parameters (The contact center may be an in-house facility to a business or enterprise for serving the enterprise in performing the functions of sales and service relative to the products and services available through the enterprise. In another aspect, the contact center may be operated by a third-party service provider, see par. [0055]);
receiving confirmation of the service request processing (when a customer indicates that he or she would like to reset their password, the agent may respond with the newly reset password, or the agent may confirm the customer's username first (without being asked), see par. [0145]);
and storing the confirmation details in the conversation context variables ( in response to a customer request to reset a password, the chatbot may be designed to always (e.g., deterministically) directly provide the newly reset password, see par. [0146]).
It would have been obvious to one of ordinary skill in the art to combine the Le invention with the teachings of Mazza for the benefit of allowing the human designer to approve, reject, or edit the suggestions, and to control the final generation of the chatbot, see par. [0074].
Regarding claim 17 Le teaches the system of claim 12, wherein the instructions further cause the system to: detecting when execution of the system action fails (machine learning model 202 may update its configurations (e.g., weights, biases, or other parameters) based on the assessment of its prediction (e.g., outputs 206) and reference feedback information (e.g., user indication of accuracy, see par. [0042]);
generating an alternative response without using the system action (Updates to the connection weights may, for example, be reflective of the magnitude of error propagated backward after a forward pass has been completed. In this way, for example, the machine learning model 402 may be trained to generate better predictions., see par. [0042]).
However Le does not teach storing information about the failed system action attempt;
and initiating a handoff to a human agent when system action failures exceed a threshold.
IN the same field of endeavor Mazza teaches The interactions may include a plurality of excluded interactions, each of the excluded interactions failing to satisfy the threshold semantic distance, see par. [0019]. Also The contact center system may further include a chatbot service server 170 configured to provide chatbot services to the contact center. FIG. 2 is a block diagram illustrating a chatbot service 170 according to one embodiment of the present invention. The chatbot service server may include a data extraction module 172 configured to extract information from transcripts of prior chat interactions (e.g., stored in the mass storage device 126 and/or the call recording server 158). The extracted data may be provided to a chatbot generation module 174, which is configured to generate a chatbot from the extracted data. The chatbot service server 170 may provide a user interface 176 for human supervisors or designers to control the configuration of the chatbots generated by the chatbot generation module 174, such as by displaying one or more suggestions extracted by the data extraction module 172, allowing the human designer to approve, reject, or edit the suggestions, and to control the final generation of the chatbot, see par. [0074].
It would have been obvious to one of ordinary skill in the art to combine the Le invention with the teachings of Mazza for the benefit of allowing the human designer to approve, reject, or edit the suggestions, and to control the final generation of the chatbot, see par. [0074].
Regarding claim 20 Le teaches the system of claim 12, wherein the conversation context variables comprise one or more of: a conversation identifier uniquely identifying the conversation (conversation details, see par. [0027]);
However Le does not teach turn numbers tracking a sequence of message exchanges; system action results organized by turn number; user intent information from the first generative language model; and parameter values extracted from user messages and system action results.
IN the same field of endeavor Mazza teaches The dialog graph contains many possible flows, mainly due to variability between the agents: the agent may or may not ask for the customer's inquiry reason; based on that reason, he may or may not choose to offer a phone call; and finally, he decides whether to up-sell another voice service product. The pruned dialogue tree organizes the agent's prompts in a consistent order, see par. [0148]. In some embodiments, the intent classification the sentences may be clustered using supervised intent classification using training examples generated by a chatbot designer (e.g., manually or selected from a list of examples), see par. [0128]. he clustering results are projected on the original chats: each sentence is replaced by its cluster's representative (e.g., the most frequently-appearing sentence in the cluster). The sentences occurring in at least a threshold percentage of the interactions are set as the anchors of the topic., see par. [0129].
It would have been obvious to one of ordinary skill in the art to combine the Le invention with the teachings of Mazza for the benefit of allowing the human designer to approve, reject, or edit the suggestions, and to control the final generation of the chatbot, see par. [0074].
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure. Pertinent prior art available on form 892.
Jonnalagadda ‘247 teaches classifying the sentence level intents uses a recurrent neural network (RNN), and paragraph level intents uses a different RNN. These recurrent neural networks use Bi-directional Long Short Term Memory (Bi-LSTM), see par. [0012].
Fincun ‘925 teaches a chatbot application for providing contextual help recommendations based on conversational context and interaction patterns, see abstract.
Gupta ‘665 teaches methods, systems, and programming for generating a conversational agent by automatic paraphrase generation based on machine translation., see abstract.
Kaur ‘948 teaches automated intent mining system extracts intents from stored conversations in a text-based chat corpus for use in an AI-driven dialog system, see abstract.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Michael Ortiz-Sanchez whose telephone number is (571)270-3711. The examiner can normally be reached Monday- Friday 9AM-6PM.
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, Bhavesh Mehta can be reached at 571-272-7453. 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.
/MICHAEL ORTIZ-SANCHEZ/Primary Examiner, Art Unit 2656
