DETAILED ACTION
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA . This office action is in response to communication filed on 9/11/2026.
The instant application having application No. 18/228,540 filed on July 31, 2023, presents claims 1-19 for examination. The instant application does not have priority data.
Continued Examination Under 37 CFR 1.114
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and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114.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 September 11, 2026 has been entered.
Status of the Claims
Claims 1 and 12 are amended, claims 1-19 are currently pending in the application.
Response to Amendment
Regarding claim objections: Applicant amendments to claims appropriately addressed the objections to the claims, the objections are withdrawn.
Regarding 35 U.S.C. § 101 rejection: Amended claims are still abstract idea without significantly more. New grounds of 101 abstract idea rejections are presented in the office action below.
Examiner Notes
Examiner cites particular columns, paragraphs, figures and line numbers in the references as applied to the claims below for the convenience of the applicant. Although the specified citations are representative of the teachings in the art and are applied to the specific limitations within the individual claim, other passages and figures may apply as well. It is respectfully requested that, in preparing responses, the applicant fully consider the references in their entirety as potentially teaching all or part of the claimed invention, as well as the context of the passage as taught by the prior art or disclosed by the examiner.
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.
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1-19 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
With respect to claim 1, This claim is within at least one of the four categories of patent eligible subject matter as it is directed to a system claim under Step 1.
Under Prong 1, Step 2A:
However, the limitations of claim 1,
“deriving, […], feature information associated with the application data and the update data, wherein the application data comprises a plurality of code sections each having one or more code blocks and the update data defines a modification to at least one code block, wherein deriving the feature information comprises:
deriving connectivity feature information defining a connectivity of code blocks to a particular code section of the plurality of code sections that will be modified by update data;
deriving historical feature information defining a frequency of updates to the particular code section of the plurality of code sections that will be modified by the update data;
deriving activity feature information defining a number of times the particular code section of the plurality of code sections that will be modified by the update data is accessed in a time period,
combining the connectivity feature information, the historical feature information, and the activity feature information to form a multi-modal vector;
predicting, [using a neural network of the prediction system, the neural network comprising a plurality of recurrent neural network layers trained with a machine learning algorithm using training data generated by pushing test updates to a replica of the application data running in a sandboxed environment, the test updates being configured to crash the application data running in the sandboxed environment], a failure probability for the update data that modifies the particular code section using the multi-modal vector;
and
initiating, responsive to the indication signifying the failure, an automated rollback procedure that includes:
identifying, using the connectivity feature information, one or more downstream code sections that depend from the particular code section;
reverting the one or more downstream code sections and the particular code section to a respective prior state that existed prior to application of the update data; and
generating […] a rollback event log associated with the automated rollback procedure.”
as drafted, are functions that, under its broadest reasonable interpretation, recite the abstract idea of a mental process. The limitations encompass a human mind carrying out the functions through observation, evaluation, judgment and /or opinion, or even with the aid of pen and paper. E.g. human can manually derive feature information as defined in the claim, can manually derive connectivity feature information as defined in the claim, can manually derive historical feature information as defined in the claim, can manually derive activity information as defined in the claim, can manually combine the derived information to form a multi-modal vector, can manually predict a failure probability for the update data as defined in the claim, can manually train the neural network as defined in the claim, can manually configure the test updates to crash the application data as defined in the claim, can manually initiate an automated rollback procedure as defined in the claim, can manually identify downstream code sections as defined in the claim, can manually revert the downstream code sections and the particular code section to a respective prior state as defined in the claim, can manually generate a rollback event log as defined in the claim. Thus, these limitations recite and fall within the “Mental Processes” grouping of abstract ideas under Prong 1 Step 2A.
Under Prong 2, Step 2A:
The judicial exception is not integrated into a practical application. The claim recites the following additional elements
“A system”, “one or more processors”, “a memory”, “a prediction system”, “a server”, “a cloud computing system”, “a production environment”, “a neural network of the prediction system, the neural network comprising a plurality of recurrent neural network layers trained with a machine learning algorithm using training data generated by pushing test updates to a replica of the application data running in a sandboxed environment,”;
“operating a prediction system on a server or cloud computing system that is independent from a production environment executing application data and update data;”
“outputting an indication signifying a failure based on the failure probability satisfying a threshold;” and
“… storing in the memory a rollback event log associated with the automated rollback procedure”
These additional elements “A system”, “one or more processors”, “a memory”, “a prediction system”, “a server”, “a cloud computing system”, “a production environment”, “a neural network comprising a plurality of recurrent neural network layers trained with a machine learning algorithm”; are cited as generic computer/program components, or merely as a tool to implement the identified abstract idea, do not integrate the judicial exception into a practical application. Refer to MPEP 2106.05(f). “pushing test updates to a replica of the application data running in a sandboxed environment” is insignificant extra-solution activity such as transmitting and storing data. The “operating …” limitation is merely using a computer (a server or cloud computing system) to implement the identified abstract idea, e.g. writing the identified mental processes into a computer program as a prediction system, and operating the prediction system on a computer. The “outputting …” limitation is insignificant extra-solution activity like transmitting data. Refer to MPEP 2106.05(g). The “storing …” limitation is insignificant extra-solution activity such as storing and retrieving information in memory, Refer to MPEP 2106.05(d) II.
Under Step 2B:
The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception. As discussed above with respect to integration of the abstract idea into a practical application, the additional elements “A system”, “one or more processors”, “a memory”, “a prediction system”, “a server”, “a cloud computing system”, “a production environment”, “a neural network comprising a plurality of recurrent neural network layers trained with a machine learning algorithm” are mere use of generic computer to implement the abstract idea, do not amount to significantly more than the judicial exception, thus, are not an inventive concept. “pushing test updates to a replica of the application data running in a sandboxed environment” is insignificant extra-solution activity such as transmitting and storing data which are recognized as well-understood, routine, conventional activity, Refer to MPEP 2106.05(d) II. The “operating …” limitation is merely using a computer (a server or cloud computing system) to implement the identified abstract idea, e.g. writing the identified mental processes into a computer program as a prediction system, and operating the prediction system on a computer. Thus, it is not an inventive concept. The “outputting …” limitation is insignificant extra-solution activity like transmitting data which is recognized as well-understood, routine, conventional activity, see MPEP 2106.05(d) II, Symantec for receiving and transmitting data. The “storing …” limitation is insignificant extra-solution activity such as storing and retrieving information in memory which is recognized as well-understood, routine, conventional activity, Refer to MPEP 2106.05(d) II, Versata Dev. Group, Inc. v. SAP Am., Inc. for retrieving and storing data. Accordingly, even viewed as a whole, the claim does not appear to be patent eligible under 35 USC 101.
With respect to claim 12, This claim is within at least one of the four categories of patent eligible subject matter as it is directed to a method claim under Step 1.
This claim recites a method that is disclosed in claim 1 and therefore recites the same abstract idea as claim 1, please see the office action analysis regarding claim 1.
Claim 12 does not recite any additional element that is not recited in claim 1.
With respect to claims 2 and 13, “wherein deriving the feature information further comprises: determining a size of the particular code section that will be modified by the update data and outputting at least the size of the particular code section that will be modified by the update data as complexity feature information and the neural network is additionally trained using the complexity feature information.” The claim as drafted, is function that, under its broadest reasonable interpretation, recites the abstract idea of a mental process. e.g. human can manually determine a size of the particular code section as defined in the claim; the “outputting …” limitation is insignificant extra-solution activity like transmitting data which is recognized as well-understood, routine, conventional activity, see MPEP 2106.05(d). Training the neural network using the complexity feature information can be performed by human manually.
With respect to claims 3 and 14, “wherein deriving the feature information further comprises: determining a trust score for the update data and output at least a trust score for the update data as social feature information wherein the trust score includes at least a number of other successful code section changes made by a person that created the update data and, wherein the neural network is additionally trained using the trust score.” The claim as drafted, is function that, under its broadest reasonable interpretation, recites the abstract idea of a mental process. e.g. human can manually determine a trust score for the update data as defined in the claim; the “output …” limitation is insignificant extra-solution activity like transmitting data which is recognized as well-understood, routine, conventional activity, see MPEP 2106.05(d). Training the neural network using the trust score can be performed by human manually.
With respect to claims 4 and 15, “wherein the trust score further includes a tenure of the person that created the update data.” The claim as drafted, is function that, under its broadest reasonable interpretation, recites the abstract idea of a mental process. e.g. human can manually calculate the trust score including consideration of a tenure of the person that created the update data.
With respect to claims 5 and 16, “wherein the trust score includes a score for rank or title within an organization for the person that created the update data.” The claim as drafted, is function that, under its broadest reasonable interpretation, recites the abstract idea of a mental process. e.g. human can manually calculate the trust score including consideration of rank or title of the person that created the update data.
With respect to claims 6 and 17, “wherein deriving the feature information further comprises: determining a sentiment of code comments in the particular code section that will be modified by the update data and outputting at least a score for sentiment as comment feature information wherein the neural network is additionally trained using the score for sentiment.” The claim as drafted, is function that, under its broadest reasonable interpretation, recites the abstract idea of a mental process. e.g. human can manually determine the sentiment of code comments as defined by the claim language; the “outputting …” limitation is insignificant extra-solution activity like transmitting data which is recognized as well-understood, routine, conventional activity, see MPEP 2106.05(d). Training the neural network using the score for sentiment can be performed by human manually.
With respect to claims 7 and 18, “further comprising a sentiment analysis neural network trained with a machine learning algorithm to determine sentiment from text strings in code comments.” The claim as drafted, is function that, under its broadest reasonable interpretation, recites the abstract idea of a mental process. That is, other than reciting “a sentiment analysis neural network trained with a machine learning algorithm”, nothing in the claim element precludes the step from practically being performed in the mind. For example, but for the “a sentiment analysis neural network trained with a machine learning algorithm” language, “determine” in the context of this claim encompasses the user manually determine sentiment from text strings in the comments of code. “a sentiment analysis neural network trained with a machine learning algorithm”; is cited as generic computer/program component, or merely as a tool to implement the identified abstract idea, does not integrate the judicial exception into a practical application, and is not an inventive concept. Refer to MPEP 2106.05(f).
With respect to claims 8 and 19, “wherein deriving the feature information further comprises determining a number of failures the particular code section that will be modified by the update data has experienced due to past updates and output the number of failures the particular code section that will be modified by the update data has experienced due to past updates as failure feature information wherein the neural network is additionally trained using the failure feature information.” The claim as drafted, is function that, under its broadest reasonable interpretation, recites the abstract idea of a mental process. e.g. human can manually determine a number of failures as defined by the claim; the “output …” limitation is insignificant extra-solution activity like transmitting data which is recognized as well-understood, routine, conventional activity, see MPEP 2106.05(d). Training the neural network using the failure feature information can be performed by human manually.
With respect to claim 9, “wherein the failure feature information further includes a number of reverts to original code.” The claim as drafted, is function that, under its broadest reasonable interpretation, recites the abstract idea of a mental process. e.g. human can manually determine whether the failure feature information further includes a number of reverts to original code.
With respect to claim 10, “wherein the particular code section includes an entire file.” The claim as drafted, is function that, under its broadest reasonable interpretation, recites the abstract idea of a mental process. e.g. human can manually determine whether the code section includes an entire file.
With respect to claim 11, “wherein the code section includes multiple files.” The claim as drafted, is function that, under its broadest reasonable interpretation, recites the abstract idea of a mental process. e.g. human can manually determine whether the code section includes multiple files.
Response to Arguments
Applicant's arguments with respect to 101 abstract idea rejections filed 9/11/2026 have been fully considered but they are not persuasive.
At p9 last to p11 first paragraph of the Remarks, Applicant argued with respect to Step 2A, Prong 2. Particularly, at p9 last to p10 first paragraph of the Remarks, Applicant argued that “The claims resolve this technological challenge by utilizing a specialized prediction system and network topology. …”
Examiner respectfully disagrees, because the prediction system and the network topology are merely using computer and software as tools to implement identified mental processes.
At p10 second paragraph of the Remarks, Applicant argued that “Furthermore, the claimed system does not rely on a generic algorithm, but specifically utilizes a recurrent neural network because its architecture "maintains historical information by feeding the result from a previous time T to a current time T+1". This specific structural configuration allows the system to process the sequential, time-based nature of the claimed "historical feature information" and "activity feature information" to accurately predict cascading failures.”
Examiner respectfully disagrees, because, maintaining historical information by feeding the result from a previous time T to a current time T+1 is mental process, the user can manually maintain this historical information. The recurrent neural network is merely used as a tool to implement the identified mental process.
At p10 third paragraph of the Remarks, Applicant argued that “Together, this specific technical configuration provides a technical improvement by actively pushing test updates into the sandbox that are configured to crash a replica system therein. By doing so, the system generates accurate, labeled failure data to train the temporal model without destroying the live production version of the system.”
Examiner respectfully disagrees, because, as set forth in the office action, configuring test updates to crash a replica system is mental process, the user can manually configure such test updates. Pushing test updates into the sandbox is insignificant extra-solution activity such as transmitting and storing data which is well understood, routine, and conventional, see MPEP 2106.05(d) II.
At p10 fourth paragraph of the Remarks, Applicant argued that “The Office Action's position that these operations constitute an abstract idea is untenable in light of recent Patent Trial and Appeal Board (PTAB) decisions related to Al inventions. In Ex parte Carmody, Appeal 2025-002843, the Board reversed a nearly identical § 101 rejection directed to Al models, holding that "the claims recite an improvement in training of models" which serves to integrate the judicial exception into a practical application.”
Examiner respectfully disagrees, because, the statement "the claims recite an improvement in training of models" is conclusive, without details it would not be sufficient to reverse the 101 rejections. For the instant case, as set forth in the office action, the training of the neural network is mental process, it does not improve technology.
At p10 fifth paragraph of the Remarks, Applicant argued that “The Board in Carmody anchored its reversal in the precedent of Ex parte Desjardins, which recognizes that "improvements in training a machine learning model" are patent-eligible technical improvements. Just as the claims in Carmody were found eligible because their specific architecture allowed models to be "updated and improved separately and independently," Applicant's claims likewise described a specific, sandbox-crashing architecture strictly for the safe training and execution of the predictive model.”
Examiner respectfully disagrees, because, Ex parte Desjardins case is not applicable here. In Desjardins case, the limitation “adjust the first values of the plurality of parameters to optimize performance of the machine learning model on the second machine learning task while protecting performance of the machine learning model on the first machine learning task” reflected the improvement disclosed in the specification. The instant claims do not recite any similar features. As set forth in the office action and as explained above, the training of the neural network is mental process, it does not improve technology.
At p10 sixth paragraph of the Remarks, Applicant argued that “Furthermore, the Board in Carmody explicitly reminded examiners of the Federal Circuit's holding in Enfish that "[s]oftware can make non-abstract improvements to computer technology, just as hardware improvements can," while stressing "the importance of recognizing patent eligibility for AI-related inventions when appropriate”.
Examiner respectfully disagrees, because, the current 101 rejections are not based on the argument “software cannot make an improvement to computer technology”. In fact, Examiner agrees that software can make an improvement to computer technology.
At p11 first paragraph of the Remarks, Applicant argued that “By claiming a specific, technical architectural improvement to how a machine-learning model is trained purposely crashing a replica environment to safeguard a production environment-the claims do not recite a mere abstract mental process. Rather, they recite a tangible, patent-eligible improvement to computer functionality.”
Examiner respectfully disagrees, because, as explained above, configuring test updates to crash a replica environment is mental process, the user can configure such test updates. Training the machine-learning model is also mental process with additional insignificant extra-solution activities such as pushing test updates into the sandbox. Thus, the claims do not improve computer functionality.
At p11 second paragraph of the Remarks, Applicant argued that “Furthermore, Applicant submits that the claims recite an ordered combination of elements that amounts to significantly more than the alleged abstract idea under Step 2B. The claimed operations cannot be performed in the human mind, nor are they a mere generic computer implementation. The claims require predicting a failure probability using a multi-modal vector via a recurrent neural network uniquely trained using failure data generated by crashing a replica of the application data in a sandboxed environment, and subsequently utilizing that model to automatically initiate a rollback procedure that physically reverts downstream code sections to a prior state.”
Examiner respectfully disagrees, because, as set forth in the office action, the claims recite mental processes and additional elements. But the additional elements are either insignificant extra-solution activities which are recognized as well understood, routine, and conventional or merely using computer and software as tools to implement the identified mental processes. Thus, even viewed as a whole, the claims do not amounts to significantly more than the judicial exception itself.
At p11 third paragraph of the Remarks, Applicant argued that “Applicant submits that utilizing a specifically trained neural network to process complex structural connectivity data and physically trigger automated computing infrastructure rollbacks constitutes an unconventional ordered combination of elements. The structural limitation of generating the training data by pushing crash-configured updates to a replica sandbox directly ties the claim to a specialized computer architecture, foreclosing any preemption of generic risk- modeling. This combination represents an inventive concept that proactively safeguards live production systems from volatile code updates, far exceeding well-understood, routine, or conventional activity.”
Examiner respectfully disagrees, because, as set forth in the office action, and as explained above, training the neural network is mental process, rollback procedure is also mental process; pushing crash-configured updates to a replica sandbox is insignificant extra-solution activity such as transmitting and storing data. These processes do not affect computer architecture. Thus, even viewed as a whole, the claims do not represent any inventive concept.
At p11 fourth paragraph of the Remarks, Applicant argued that “For at least these reasons, the Applicant respectfully requests withdrawal of the § 101 rejection of the pending claims.”
Examiner respectfully disagrees, because, as set forth in the office action, and as explained above, the claims do not appear to be patent eligible under 35 USC 101.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure. For example, Tarlano et al. US 20190102244 A1 teaches method and system for predicting failure events.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Zengpu Wei whose telephone number is 571-270-1302. The examiner can normally be reached on Monday to Friday from 8:00AM to 5:00 PM.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Bradley Teets, can be reached on 571-272-3338. 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). If you would like assistance from a USPTO Customer Service Representative or access to the automated information system, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/ZENGPU WEI/Examiner, Art Unit 2197
