DETAILED ACTION
Notice of Pre-AIA or AIA Status
01. The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Continued Examination Under 37 CFR 1.114
02. A request for continued examination under 37 CFR 1.114, including the fee set forth in 37 CFR 1.17(e), was filed in this application after final rejection. Since this application is eligible for continued examination under 37 CFR 1.114, and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114. Applicant's submission filed on 03/10/2026 has been entered.
Response to Arguments
03. Applicant’s arguments with respect to claims 1, 5 – 13, and 17 – 26 have been considered, but are moot in view of the new ground(s) of rejection.
Applicant argues that not all claimed limitations in claim 1 are disclosed by Asokan, in that Asokan does not teach providing feedback by a first trained machine learning tool. Examiner respectfully agrees. However, a new prior art reference has been relied upon to teach this claimed limitation. The same holds true claims 12 and 20, which have similar arguments presented.
Applicant argues that not all claimed limitations in claim 19 are disclosed by Asokan, in that Asokan does not teach less than 160 billion parameters. Examiner respectfully agrees. However, a new prior art references has been relied upon to teach this claim limitation. The same holds true for claim 9, which has similar arguments presented.
Applicant argues that not all claimed limitations in claim 12 are disclosed by Asokan and Chen. More specifically, Applicant argues that Chan does not teach “causing the trained third ML tool to be deployed to perform at least in part of the emulated workflow”. Examiner respectfully disagrees. The main argument presented by Applicant is that the Chan the only machine learning application is improving the code generation process. However, any of the described steps may be performed by machine learning applications. For example, Applicant refers to software entities in Chan that perform workflow, dynamically generate code, improve the code generation, etc. It is unclear to Examiner why only the code generation process could be reasonably taught as utilizing machine learning, as the machine learning could be applied to any of these software entities. Furthermore, the claim is rejected under a combine teachings of Chan with Asokan, so the modified teaching of Asokan would incorporate the taught features of Chan. Asokan teaches various different machine learning models that perform different functionality, such as machine learning models that utilize trait scores [0012], analyze responses [0014], performs resume profile ranking [0072], and generates questions [0075]. Therefore, Asokan makes it clear that various machine learning models are developed and deployed that can be done to perform any of the functionalities that could be incorporated, including the emulated workflow taught by Chan. Therefore the prior art is understood to teach the Applicant’s aforementioned claim limitations.
Applicant argues that not all claimed limitations in claim 17 are disclosed by Rose. Examiner respectfully agrees, and claim 17 has been indicated as being allowable.
Applicant argues that not all claimed limitations in claim 24 are disclosed by Asokan and Focke. Examiner respectfully disagrees. Claim 24 has been amended to recite that the workflow map evolves over the temporal course of an associated interview representation. Although this further defines the workflow over simply being “dynamic”, this claim limitation is understood by Examiner to simply mean that the workflow is based off the interview. The claim already recites that the workflow map is based off the training data records, so as more training data records become available, i.e. by continuing the interview, the workflow map would then include those additional training data records. In other words, the claim only suggests or recites that the additional data would update the structure that includes that data, which would be inherent.
Claim Rejections - 35 USC § 103
04. In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
05. 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 of this title, 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.
06. Claims 1, 10, 11, and 20 are rejected under 35 U.S.C. 103 as being unpatentable over Asokan et al. (US PGPub 2021/0312399), hereinafter “Asokan”, in view of Janz (US PGPub 2018/0025303).
Consider claim 1, Asokan discloses a method, comprising:
by a first trained machine learning (ML) tool (paragraph [0038], a trained machine learning model is used to perform different actions);
monitoring an interview of one or more skilled personnel interviewee(s) conducted by one or more interviewers including a second trained ML tool distinct from the first trained ML tool (paragraphs [0002], [0015], an interview is conducted between two parties, such as an automated chat resource and a candidate. There are any number of different machine learning models that are used, such as a machine learning model that utilizes trait scores [0012], a machine learning model that analyzes responses [0014], a machine learning model that does resume profile ranking [0072], a machine learning model that generates questions [0075], etc.);
providing, after the interview, feedback for at least the second ML tool based on the monitoring (paragraph [0038], as a result of conducting an interview, feedback can be generated based on the interactions and responses given);
creating a plurality of training data records based on the feedback (paragraphs [0012], [0038], [0084], a machine learning model is trained based on data that is obtained by way of conducting an interview, which can include different types of data that is used in the training steps);
causing updating of the second trained ML tool by reinforcement learning using at least some of the training data records (paragraphs [0047], [0063], a machine learning model is updated, which includes the model learning what types of questions to answer as a result of performing and processing the interviews. This would lead to a type of reinforcement learning as the model would become more humanistic and learn to ask better questions during the interviews).
However, Asokan does not specifically teach that the feedback is provided by the machine learning tool, but is instead limited to manual feedback providing by an entity.
In the same field of endeavor, Janz discloses a method comprising:
by a first trained learning machine (ML) tool: providing, after the interview, feedback (paragraphs [0174], [0219], [0225], feedback is generated and provided to the system based on an automated process).
Therefore it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the automated feedback process for interviews taught by Janz into the automated interviewing taught by Asokan for the purpose of allowing the feedback generation and incorporation process to be automated and without user intervention, which would lead to faster and more accurate feedback that can then be incorporate to make the model work better in the future.
Consider claim 10, and as applied to claim 1 above, Asokan discloses a method comprising:
the one or more interviewers are a single interviewer which is the second trained ML tool (paragraphs [0002], [0015], [0038], [0050], the interviewer is an automated chat resource, which can be part of a machine learning model).
Consider claim 11, and as applied to claim 1 above, Asokan discloses a method comprising:
the one or more skilled personnel are a single skilled person (paragraphs [0002], [0049], the candidate, which is being interviewed, can be a single candidate/person).
Claim 20 recites the same embodiments as those found in claim 1. The only difference is that either a method or computer-readable media is being claimed. Since the same claim limitations are otherwise present, claim 20 is rejected under the same rationale as that provided for claim 1.
07. Claims 5, 6, 22, and 24 are rejected under 35 U.S.C. 103 as being unpatentable over Asokan et al. (US PGPub 2021/0312399), hereinafter “Asokan”, in view of Janz (US PGPub 2018/0025303), in further view of Chow (US PGPub 2022/0198298), hereinafter “Chow”.
Consider claim 5, and as applied to claim 1 above, Asokan and Janz disclose the claimed invention except for a workflow that describes a job function.
In the same field of endeavor, Chow discloses a method comprising:
at least a given one of the created training data record(s) comprises, at least in part, a static or dynamic first workflow map describing a job function performed by the one or more skilled personnel interviewee(s) (paragraphs [0030], [0031], workflows are recorded, which can be associated with attributes such as a job function).
Therefore it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the workflow that incorporates a job function taught by Chow into the automated interviewing taught by Asokan and Janz for the purpose of allowing additional types of data to be used to train the machine learning model so that the model could be enhanced by being able to ask better questions and solicit better responses from candidates during the interview process. This would lead to a better understanding of the candidates and improve the process of hiring people for jobs.
Consider claim 6, and as applied to claim 5 above, Asokan discloses a method comprising: the updating improves a function of the second trained ML tool to generate second workflow maps for respective interviews (paragraphs [0038], [0041], [0047], [0075], the training of the machine learning models allows for the models to have improved functionality, such as being able to ask better questions, assess candidates better, etc.).
Claim 22 recites the same embodiments as those found in claim 5. The only difference is that either a method or computer-readable media is being claimed. Since the same claim limitations are otherwise present, claim 22 is rejected under the same rationale as that provided for claim 5.
Consider claim 24, and as applied to claim 22 above, Asokan discloses a method comprising:
the first workflow map evolves over the temporal course of an associated interview representation (paragraphs [0038], [0041], [0047], [0075], data in continually received throughout the duration of the interview, such that the additional data that is received is added to a workflow).
08. Claims 7, 8, and 18 are rejected under 35 U.S.C. 103 as being unpatentable over Asokan et al. (US PGPub 2021/0312399), hereinafter “Asokan”, in view of Janz (US PGPub 2018/0025303), in further view of Rose (US PGPub 2021/0174347).
Consider claim 7, and as applied to claim 1 above, Asokan and Janz disclose the claimed invention except that microservices are used.
In the same field of endeavor, Rose discloses a method comprising:
the first trained ML tool is a copilot comprising network of microservices including an expansion microservice, a retrieval microservice, one or more core microservices, and one or more evaluation microservices (paragraphs [0092] – [0094], a network of microservices are used, such that any amount of microservices can be used, which all have a different type of processing capability and can service a different purpose).
Therefore it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the microservices architecture taught by Rose into the automated interviewing taught by Asokan for the purpose of being able to use different types of computing infrastructure to be able to conduct the interviews, which would allow for additional capabilities for the interviews through more advanced monitoring techniques.
Consider claim 8, and as applied to claim 1 above, Asokan and Janz disclose the claimed invention except that a deep neural network is used.
In the same field of endeavor, Rose discloses a method comprising:
the first trained ML tool comprises at least one large language model (LMLM) or at least one deep neural network (DNN) (paragraphs [0501], [0502], a deep neural network is used for the machine learning model).
Therefore it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the deep neural network for a machine learning model taught by Rose into the automated interviewing taught by Asokan for the purpose of being able to utilize different and more advanced models for the interviewing process, which would lead to being able to ask better questions during the interview and better process the results of the interview.
Claim 18 recites the same embodiments as those found in claim 8. The only difference is that either a method or computer-readable media is being claimed. Since the same claim limitations are otherwise present, claim 18 is rejected under the same rationale as that provided for claim 8.
09. Claims 9 and 19 are rejected under 35 U.S.C. 103 as being unpatentable over Asokan et al. (US PGPub 2021/0312399), hereinafter “Asokan”, in view of Janz (US PGPub 2018/0025303), in further view of Tomut et al. (US PGPub 2025/0111212), hereinafter “Tomut”.
Consider claim 9, and as applied to claim 1 above, Asokan and Janz disclose the claimed invention except that the machine learning tool has less than 16 billion parameters.
In the same field of endeavor, Tomut discloses a method comprising:
the first trained ML tool has less than 160 billion parameters (paragraphs [0011], [0046], [0096], the number of parameters that are used for a machine learning model can be 7 billion or 2 billion, which is less than 8 billion).
Therefore it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the number of parameters taught by Tomut into the automated interviewing taught by Asokan and Janz for the purpose of different sized models to be used, which can have different levels of training performed on them and therefore have different capabilities.
Consider claim 19, and as applied to claim 20 above, Asokan and Janz disclose the claimed invention except that the machine learning tool has less than 16 billion parameters.
In the same field of endeavor, Tomut discloses a method comprising:
the first trained ML tool has less than 160 billion parameters (paragraphs [0011], [0046], [0096], the number of parameters that are used for a machine learning model can be 7 billion or 2 billion, which is less than 8 billion) and is implemented on a single compute node having at most eight GPUs, each GPU being an accelerator or coprocessor (paragraphs [0010], [0011], [0042], [0050], GPUs are used, such that a single GPU can be used for the processing).
Therefore it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the number of parameters and GPUs taught by Tomut into the automated interviewing taught by Asokan and Janz for the purpose of different sized models to be used, which can have different levels of training performed on them and therefore have different capabilities.
10. Claims 12, 13, and 25 are rejected under 35 U.S.C. 103 as being unpatentable over Asokan et al. (US PGPub 2021/0312399), hereinafter “Asokan”, in view of Janz (US PGPub 2018/0025303), in further view of Chan et al. (US PGPub 2024/0005242), hereinafter “Chan”.
Consider claim 12, Asokan discloses a method, comprising:
by a first trained machine learning (ML) tool (paragraph [0038], a trained machine learning model is used to perform different actions);
monitoring an interview of one or more skilled personnel interviewee(s) conducted by one or more interviewers including a second trained ML tool distinct from the first trained ML tool (paragraphs [0002], [0015], an interview is conducted between two parties, such as an automated chat resource and a candidate. There are any number of different machine learning models that are used, such as a machine learning model that utilizes trait scores [0012], a machine learning model that analyzes responses [0014], a machine learning model that does resume profile ranking [0072], a machine learning model that generates questions [0075], etc.);
providing, after the interview, feedback for at least the second ML tool based on the monitoring (paragraph [0038], as a result of conducting an interview, feedback can be generated based on the interactions and responses given);
creating a plurality of training data records based on the feedback (paragraphs [0012], [0038], [0084], a machine learning model is trained based on data that is obtained by way of conducting an interview, which can include different types of data that is used in the training steps);
causing updating of the second trained ML tool by reinforcement learning using at least some of the training data records (paragraphs [0047], [0063], a machine learning model is updated, which includes the model learning what types of questions to answer as a result of performing and processing the interviews. This would lead to a type of reinforcement learning as the model would become more humanistic and learn to ask better questions during the interviews);
generating additional training data from the monitored interview (paragraphs [0047], [0075], [0084], [0089], any amount of training can be performed, which can include obtaining sets of training data from different sources, such as from responses, sources, feedback, etc.);
using the additional training data to train a third ML tool to (paragraphs [0047], [0075], [0084], [0089], the training data is used to train one or more machine learning models).
However, Asokan does not specifically teach that the feedback is provided by the machine learning tool, but is instead limited to manual feedback providing by an entity.
In the same field of endeavor, Janz discloses a method comprising:
by a first trained learning machine (ML) tool: providing feedback after the interview (paragraphs [0174], [0219], [0225], feedback is generated and provided to the system based on an automated process).
Therefore it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the automated feedback process for interviews taught by Janz into the automated interviewing taught by Asokan for the purpose of allowing the feedback generation and incorporation process to be automated and without user intervention, which would lead to faster and more accurate feedback that can then be incorporate to make the model work better in the future.
However, Asokan and Janz do not specifically disclose that the machine learning model can emulate workflow of the interviewee.
In the same field of endeavor, Chan discloses a method comprising:
emulate workflow performed by at least one of the skilled personnel interviewee(s) (paragraphs [0013], [0042], workflow modeling is used to automate actions that someone performs, such as in a job or work environment);
causing the trained third ML tool to be deployed to perform at least part of the emulated workflow (paragraphs [0013], [0042], [0043], machine learning techniques are applied to allow for the automation of the workflow).
Therefore it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the emulation of workflow taught by Chan into the automated interviewing taught by Asokan and Janz for the purpose of allowing the functions that a person performs, such as with respect to their job and job responsibilities, to be performed and automated by a system, which would lead to improved job functionality and a reduced need on humans to perform tasks.
Consider claim 13, and as applied to claim 12 above, Asokan discloses a method comprising:
the one or more interviewers are one or more first interviewers (paragraph [0002], an automated interview session is performed) and the method further comprises, prior to the monitoring:
training a trainee ML tool to perform a next prompt prediction task, using a training corpus comprising respective representations of one or more training interviews (paragraphs [0038], [0084], the machine learning model is used to make predictions), wherein each training interview comprises an alternating sequence of (i) prompts from one or more second interviewers to one or more subjects and (ii) responses from the one or more subjects (paragraphs [0047], [0089], an automated interview sessions is performed, which includes questions and answers that are given by the two parties during the interview);
subsequently deploying an instance of the trainee ML tool as the first trained ML tool (paragraphs [0038], [0049], [0050], a machine learning model is created that is used for the automated recruitment system).
Consider claim 25, and as applied to claim 12 above, Chan discloses a method comprising:
the additional training data comprises at least one workflow map (paragraphs [0015], [0023], workflows are defined that are used in the machine learning techniques).
12. Claims 21 and 23 are rejected under 35 U.S.C. 103 as being unpatentable over Asokan et al. (US PGPub 2021/0312399), hereinafter “Asokan”, in view of Janz (US PGPub 2018/0025303), in further view of Chow (US PGPub 2022/0198298), hereinafter “Chow”, in further view of Focke (US PGPub 2025/0045848), hereinafter “Focke”.
Consider claim 21, and as applied to claim 5 above, Asokan, Janz, and Chow disclose the claimed invention except that the workflow map is a flowchart.
In the same field of endeavor, Focke discloses a method comprising:
the first workflow map comprises a flowchart (paragraph [0022], a workflow can be defined as a flow-chart).
Therefore it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the flowchart taught by Focke into the automated interviewing taught by Asokan, Janz, and Chow for the purpose of allowing additional types of workflows to be used so that the system could be expanded to incorporate a wider range of activities that a person performs.
Consider claim 23, and as applied to claim 22 above, Asokan, Janz, and Chow disclose the claimed invention except that the workflow map is static.
In the same field of endeavor, Focke discloses a method comprising:
the first workflow map is a static workflow map (paragraph [0089], the workflow is a static workflow).
Therefore it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the static workflow taught by Focke into the automated interviewing taught by Asokan, Janz, and Chow for the purpose of allowing additional types of workflows to be used so that the system could be expanded to incorporate a wider range of activities that a person performs.
13. Claim 26 is rejected under 35 U.S.C. 103 as being unpatentable over Asokan et al. (US PGPub 2021/0312399), hereinafter “Asokan”, in view of Janz (US PGPub 2018/0025303), in further view of Chan et al. (US PGPub 2024/0005242), hereinafter “Chan”, in further view of Focke (US PGPub 2025/0045848), hereinafter “Focke”.
Consider claim 26, and as applied to claim 25 above, Asokan, Janz, and Chan disclose the claimed invention except that the workflow map is a flowchart.
In the same field of endeavor, Focke discloses a method comprising:
the first workflow map comprises a flowchart (paragraph [0022], a workflow can be defined as a flow-chart).
Therefore it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the flowchart taught by Focke into the automated interviewing taught by Asokan, Janz, and Chan for the purpose of allowing additional types of workflows to be used so that the system could be expanded to incorporate a wider range of activities that a person performs.
Allowable Subject Matter
14. Claim 17 is objected to as being dependent upon a rejected base claim, but would be allowable if rewritten in independent form, including all of the limitations of the base claim and any intervening claims.
Reasons for the Indication of Allowable Subject Matter
15. The following is a statement of reasons for the indication of allowable subject matter:
The primary reason for allowance of claim 17 in the instant application is because the prior arts of record do not teach or suggest client input going to an expansion microservice, that then goes to a retrieval microservice, that then goes to a core microservices, and then goes to an evaluation microservice, and is then conditionally directed towards the client. The prior art of record including the disclosures above neither anticipates nor renders obvious the above recited combination.
As allowable subject matter has been indicated, applicant's reply must either comply with all formal requirements or specifically traverse each requirement not complied with. See 37 CFR 1.111(b) and MPEP § 707.07(a).
Conclusion
16. Any inquiry concerning this communication or earlier communications from the Examiner should be directed to Christopher Raab whose telephone number is (571) 270-1090. The Examiner can normally be reached on Monday-Friday from 9:00am to 5:00pm.
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, Ajay Bhatia can be reached on (571) 272-3906. The fax phone number for the organization where this application or proceeding is assigned is (571) 273-8300.
Information regarding the status of an application may be obtained from the Patent Application Information Retrieval (PAIR) system. Status information for published applications may be obtained from either Private PAIR or Public PAIR. Status information for unpublished applications is available through Private PAIR only. For more information about the PAIR system, see http://pair-direct.uspto.gov. Should you have questions on access to the Private PAIR system, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free) or 703-305-3028.
/CHRISTOPHER J RAAB/Primary Examiner, Art Unit 2156
March 21, 2026