DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, 18/627,424 filed 04/04/2024, claims priority from U.S. Provisional Application 63/494,224, filed 04/04/2023.
The effective filing date is after the AIA date of March 16, 2013, and so the application is being examined under the “first inventor to file” provisions of the AIA .
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.
Status of the Application
This Second-Action Non-Final Office Action is in response to Applicant’s communication of Sept. 8, 2025.
Claims 1-15 are pending, of which claims 1, 9, and 11 are independent. Claims 6-15 are newly added. No claims have been amended or cancelled.
Claims 1-8 have been examined on the merits.
Claims 9-15 are subject to a restriction and/or election requirement.
Election/Restrictions
Newly submitted claims 9-15 are directed to an invention that is independent or distinct from the invention originally claimed for the following reasons: Both of the new independent claims 9 and 11 recite features recited neither in the previously presented independent claim 1 nor in the other of the two new independent claims. For example, new independent claim 9 recites “receiving, from the telematics device, vehicle operation data”, and new independent claim 11 recites “transmit telemetry data to the remote underwriting server through a secure application programming interface”. Neither of these features is recited in independent claim 1, and neither of these features is recited in the other of the two new independent claims. Therefore, these three independent claims recite “sub-combinations usable together”, which constitutes grounds for restriction.
Since applicant has received an action on the merits for the originally presented invention, this invention has been constructively elected by original presentation for prosecution on the merits. Accordingly, claims 9-15 are withdrawn from consideration as being directed to a non-elected invention. See 37 CFR 1.142(b) and MPEP § 821.03.
To preserve a right to petition, the reply to this action must distinctly and specifically point out supposed errors in the restriction requirement. Otherwise, the election shall be treated as a final election without traverse. Traversal must be timely. Failure to timely traverse the requirement will result in the loss of right to petition under 37 CFR 1.144. If claims are subsequently added, applicant must indicate which of the subsequently added claims are readable upon the elected invention.
Should applicant traverse on the ground that the inventions are not patentably distinct, applicant should submit evidence or identify such evidence now of record showing the inventions to be obvious variants or clearly admit on the record that this is the case. In either instance, if the examiner finds one of the inventions unpatentable over the prior art, the evidence or admission may be used in a rejection under 35 U.S.C. 103 or pre-AIA 35 U.S.C. 103(a) of the other invention.
Claim Rejections - 35 USC § 112
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
Claims 6 and 7 are rejected under 35 U.S.C. 112(b), as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor, or for pre-AIA the applicant regards as the invention.
Claim 6 and 7 recite the limitation "the telematics device". There is insufficient antecedent basis for this limitation in the claim.
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 1-8 are rejected under 35 U.S.C. §101 because the claimed invention is directed to non-statutory subject matter. The claimed invention is directed to an abstract idea, without “significantly more”.
Based on the flowchart in MPEP § 2106, Step 1 of the Alice/Mayo analysis is: “Is the claim to a process, machine, manufacture or composition of matter?”
In regards to Step 1 of the Alice/Mayo analysis, independent claim 1 is a method claim.
For the sake of compact prosecution, we continue with the Alice/Mayo “abstract idea” analysis.
Step 2A, prong 1 of the Alice/Mayo analysis is: “Does the claim recite a law of nature, a natural phenomenon (product of nature), or an abstract idea?”
In regards to Step 2A, prongs 1 and 2 of the Alice/Mayo analysis, the abstract idea elements recited in claims 1-8 are shown in italic font. (The “additional elements” and “extra solution steps” are shown in italic and underlined font):
In regards to claim 1,
1. A computer-implemented method for underwriting and pricing of loans for rideshare vehicles, the method comprising:
providing a loan pre-qualification application to a rideshare driver;
collecting credit and rental history information from the rideshare driver;
collecting information from third parties about the rideshare driver;
utilizing a base layer underwriting model to make a default risk prediction based on the information collected from the rideshare driver and the third parties;
refining the default risk prediction from the base layer underwriting model with available alternative data about the rideshare driver including the rental history and any specific business conditions;
approving or denying a loan based on the risk prediction;
optimizing the price of a loan and win probability at a desired yield by employing a price optimization and win probability model based on the refined outputs of the underwriting model, wherein the price optimization and win probability model comprises
a dense feedforward neural network with a linear bypass; and
a final layer that translates dense nodes in hidden layers in the dense feedforward neural network into parameters for a sigmoid function and feeding hold-out variables into the sigmoid function to generate a final output constrained to follow the sigmoid relationship;
determining the optimal price and optimal win probability from the final output constrained to follow the sigmoid relationship; and
utilizing a loan terms model to set loan terms that are market competitive.
In regards to claim 2,
2. The method of claim 1, further comprising
tracking parameters about the loan including payment schedule and payments made;
enabling the rideshare driver ability to make payments; and
updating the loan information regarding payments.
In regards to claim 3,
3. The method of claim 2, further comprising
installing a telematics device on the vehicle financed with the loan,
wherein the telematics device is capable of tracking location of the vehicle, communicating with the rideshare driver and disabling ignition of the vehicle in event of loan default.
In regards to claim 4,
4. The method of claim 1, wherein the loan terms include weekly payments.
In regards to claim 5,
5. The method of claim 1, wherein the collecting credit and rental history information from the rideshare driver includes the rideshare driver granting bank account access to a third-party financial services company.
In regards to claim 6,
6. (New) The method of claim 1, wherein the base underwriting model receives updated borrower-specific data from the telematics device and retrains the neural-network parameters based on the updated data to refine the price optimization output.
In regards to claim 7,
7. (New) The method of claim 1, further comprising transmitting a vehicle-status signal from the telematics device to a server executing the underwriting model, wherein the signal is processed to modify model weighting of payment-history features.
In regards to claim 8,
8. (New) The method of claim 1, wherein the underwriting model comprises a base model trained using gradient-boosting techniques and a refinement process employing Monte-Carlo sampling to generate feature distributions used as inputs to the dense neural network.
More specifically, claims 1-8 recite an abstract idea: “Certain Methods of Organizing Human Activity", specifically “Commercial or Legal Interactions (Including Agreements in the form of Contracts; Legal Obligations; Advertising, Marketing, or Sales Activities or Behaviors; Business Relations)”, as discussed in MPEP §2106(a)(2) Parts (I) and (II), and in the 2019 Revised Patent Subject Matter Eligibility Guidance.
The “Commercial or Legal Interactions” elements include:
Claim 1: “utilizing a base layer underwriting model to make a default risk prediction based on the information collected from the rideshare driver and the third parties”.
Claim 1: “refining the default risk prediction from the base layer underwriting model with available alternative data about the rideshare driver including the rental history and any specific business conditions”.
Claim 1: “approving or denying a loan based on the risk prediction”.
Claim 1: “optimizing the price of a loan and win probability at a desired yield by employing a price optimization and win probability model based on the refined outputs of the underwriting model”.
Claim 1: “determining the optimal price and optimal win probability from the final output constrained to follow the sigmoid relationship”.
Claim 1: “utilizing a loan terms model to set loan terms that are market competitive”.
Claim 2: “further comprising … enabling the rideshare driver ability to make payments”.
Claim 4: “wherein the loan terms include weekly payments”.
Claim 5: “wherein the collecting credit and rental history information from the rideshare driver includes the rideshare driver granting bank account access to a third-party financial services company”.
Moreover, claims 1-8 also recite “Mathematical Concepts", specifically “Mathematical Relationships”, “Mathematical Formulas or Equations”, and “Mathematical Calculations”, as discussed in MPEP §2106.04(a)(2) Part (IV), and in the 2019 Revised Patent Subject Matter Eligibility Guidance.
The mathematic elements in independent claim 1 include:
“wherein the price optimization and win probability model comprises ... a dense feedforward neural network with a linear bypass”.
“wherein the price optimization and win probability model comprises ... a final layer that translates dense nodes in hidden layers in the dense feedforward neural network into parameters for a sigmoid function and feeding hold-out variables into the sigmoid function to generate a final output constrained to follow the sigmoid relationship”.
In an alternative interpretation, the machine learning model (i.e. the neural network) is an additional element that is merely being run on the general purpose computer, and applied to the abstract idea (loans terms, risk of default).
The “additional elements” include:
“computer-implemented method” (Claim 1),
“a telematics device” (Claim 3),
“a server” (Claim 7),
“wherein the telematics device is capable of tracking location of the vehicle” (Claim 3),
“wherein the telematics device is capable of … disabling ignition of the vehicle in event of loan default” (Claim 3),
In the alternative interpretation, the machine learning model (i.e. the neural network) is an additional element that is merely being run on the general purpose computer, and applied to the abstract idea (loans terms, risk of default).
Moreover, “additional extra-solution elements” include:
Claim 1: “providing a loan pre-qualification application to a rideshare driver”.
Claim 1: “collecting credit and rental history information from the rideshare driver”.
Claim 1: “collecting information from third parties about the rideshare driver”.
Claim 2: “tracking parameters about the loan including payment schedule and payments made”.
Claim 2: “updating the loan information regarding payments”.
Claim 3: “wherein the telematics device is capable of … communicating with the rideshare driver”.
Step 2A, prong 2 of the Alice/Mayo analysis is “Does the claim recite additional elements that integrate elements that integrate the judicial exception into a practical application?”
In regards to Step 2A, prong 2 of the Alice/Mayo analysis, this abstract idea is not integrated into a practical application, because:
The claim is directed to an abstract idea with additional generic computer elements. The generically recited computer elements (“computer-implemented method” (Claim 1), “a telematics device” (Claim 3), “a server” (Claim 7), “wherein the telematics device is capable of tracking location of the vehicle” (Claim 3), and “wherein the telematics device is capable of … disabling ignition of the vehicle in event of loan default” (Claim 3), and in the alternative interpretation, “the machine learning model”) do not add a meaningful limitation to the abstract idea, because they amount to simply implementing the abstract idea on a computer. The claim amounts to adding the words "apply it" (or an equivalent) with the abstract idea, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea.
The extra-solution activities (In claim 1: “providing a loan pre-qualification application to a rideshare driver”, “collecting credit and rental history information from the rideshare driver”, “collecting information from third parties about the rideshare driver”. In claim 2: “tracking parameters about the loan including payment schedule and payments made”, “updating the loan information regarding payments”. In claim 3: “wherein the telematics device is capable of … communicating with the rideshare driver”) do not add a meaningful limitation to the method, as they are insignificant extra-solution activity;
The combination of the abstract idea with the additional elements (generically recited computer elements), and/or with the extra-solution activities, does not integrate the abstract idea into a practical application.
The claim merely generally links the use of the abstract idea to a particular technological environment or field of use (e.g. electric grid data is not claiming the actual electric grid)- see MPEP 2106.05(h) ].
Step 2B of the Alice/Mayo analysis is: “Does the claim recite additional elements that amount to significantly more than the judicial exception?”
In regards to Step 2B of the Alice/Mayo analysis, the claims do not include additional elements that are sufficient to amount to significantly more than the abstract idea, because:
When considering the elements "alone and in combination" (“computer-implemented method” (Claim 1), “a telematics device” (Claim 3), “a server” (Claim 7), “wherein the telematics device is capable of tracking location of the vehicle” (Claim 3), and “wherein the telematics device is capable of … disabling ignition of the vehicle in event of loan default” (Claim 3), and in the alternative interpretation, “the machine learning model”), they do not add significantly more (also known as an "inventive concept") to the exception, because they amount to simply implementing the abstract idea on a computer.
Instead, they merely add the words "apply it" (or an equivalent) with the abstract idea, or mere instructions to implement an abstract idea on a computer, or merely use a computer as a tool to perform an abstract idea.
Moreover, in regards to dependent claims 2 and 3, the claimed technological features (“tracking parameters about the loan including payment schedule and payments made”, “wherein the telematics device is capable of tracking location of the vehicle” and “wherein the telematics device is capable of … disabling ignition of the vehicle in event of loan default”) were well known, routine, and conventional features, as shown in the following articles:
"Late Payment? A ‘Kill Switch’ Can Strand You and Your Car" by E. S. Povich.
"Miss a Payment? Good Luck Moving That Car" by M. Corkery and J. Silver-Greenberg.
"Dealer Deactivated Cars With ‘Kill Switch' After Borrowers' Payments Fell Short: Suit" by NBC 5Chicago.
In regards to the extra solution activities (In claim 1: “providing a loan pre-qualification application to a rideshare driver”, “collecting credit and rental history information from the rideshare driver”, “collecting information from third parties about the rideshare driver”. In claim 2: “tracking parameters about the loan including payment schedule and payments made”, “updating the loan information regarding payments”. In claim 3: “wherein the telematics device is capable of … communicating with the rideshare driver”), these are recognized as such by the court decisions listed in MPEP § 2106.05(d).
More specifically, in regards to the “storing” step (e.g. in claim 2: “updating the loan information regarding payments”), see the court cases Versata Dev. Group, Inc. v. SAP Am., Inc., 793 F.3d 1306, 1334, 115 USPQ2d 1681, 1701 (Fed. Cir. 2015) (storing and retrieving information in memory); and OIP Techs., Inc., v. Amazon.com, Inc., 788 F.3d 1359, 1363, 115 USPQ2d 1090, 1092-93 (Fed. Cir. 2015) (storing and retrieving information in memory).
More specifically, in regards to the “receiving” and “communicating” steps (e.g. in claim 1: “providing a loan pre-qualification application to a rideshare driver”, “collecting credit and rental history information from the rideshare driver”, “collecting information from third parties about the rideshare driver”. In claim 2: “tracking parameters about the loan including payment schedule and payments made”. In claim 3: “wherein the telematics device is capable of … communicating with the rideshare driver”), see the court cases OIP Techs., Inc., v. Amazon.com, Inc., 788 F.3d 1359, 1363, 115 USPQ2d 1090, 1093 (Fed. Cir. 2015) (sending messages over a network) and (presenting offers and gathering statistics), OIP Techs., 788 F.3d at 1362-63, 115 USPQ2d at 1092-93; buySAFE, Inc. v. Google, Inc., 765 F.3d 1350, 1355, 112 USPQ2d 1093, 1096 (Fed. Cir. 2014) (computer receives and sends information over a network).
Moreover, further in regards to “apply it”, according to MPEP § 2106.05(f)(2):
Use of a computer or other machinery in its ordinary capacity for economic or other tasks (e.g., to receive, store, or transmit data) or simply adding a general purpose computer or computer components after the fact to an abstract idea (e.g., a fundamental economic practice or mathematical equation) does not integrate a judicial exception into a practical application or provide significantly more. See Affinity Labs v. DirecTV, 838 F.3d 1253, 1262, 120 USPQ2d 1201, 1207 (Fed. Cir. 2016) (cellular telephone); TLI Communications LLC v. AV Auto, LLC, 823 F.3d 607, 613, 118 USPQ2d 1744, 1748 (Fed. Cir. 2016) (computer server and telephone unit). Similarly, "claiming the improved speed or efficiency inherent with applying the abstract idea on a computer" does not integrate a judicial exception into a practical application or provide an inventive concept. Intellectual Ventures I LLC v. Capital One Bank (USA), 792 F.3d 1363, 1367, 115 USPQ2d 1636, 1639 (Fed. Cir. 2015).
In contrast, a claim that purports to improve computer capabilities or to improve an existing technology may integrate a judicial exception into a practical application or provide significantly more. McRO, Inc. v. Bandai Namco Games Am. Inc., 837 F.3d 1299, 1314-15, 120 USPQ2d 1091, 1101-02 (Fed. Cir. 2016); Enfish, LLC v. Microsoft Corp., 822 F.3d 1327, 1335-36, 118 USPQ2d 1684, 1688-89 (Fed. Cir. 2016). See MPEP §§ 2106.04(d)(1) and 2106.05(a) for a discussion of improvements to the functioning of a computer or to another technology or technical field.
The Examiner holds that the independent claims “use a computer or other machinery in its ordinary capacity for economic or other tasks (e.g., to receive, store, or transmit data)” or “simply add a general purpose computer or computer components after the fact to an abstract idea”.
Claim Rejections - 35 USC § 103
This application currently names joint inventors. In considering patentability of the claims the examiner presumes that the subject matter of the various claims was commonly owned as of the effective filing date of the claimed invention(s) absent any evidence to the contrary. Applicant is advised of the obligation under 37 CFR 1.56 to point out the inventor and effective filing dates of each claim that was not commonly owned as of the effective filing date of the later invention in order for the examiner to consider the applicability of 35 U.S.C. 102(b)(2)(C) for any potential 35 U.S.C. 102(a)(2) prior art against the later invention.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
The factual inquiries set forth in Graham v. John Deere Co., 383 U.S. 1, 148 USPQ 459 (1966), that are applied for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claims 1, 2, and 5-8 are rejected under 35 U.S.C. 103 as being unpatentable over US 2023/0260020 A1 to Kim et al. (“Kim”, Eff. Filed Jun. 17, 2019) in view of US 2023/0186385 A1 to Masson (“Masson”. Eff. Filed Dec. 13 or 14, 2021), and further in view of US 2018/0308013 A1 to O’Shea (“O’Shea”. Eff. Filed on Apr. 24, 2017. Published on Oct. 25, 2018) and in view of “Nonlinear Principal Component Analysis Using Auto-associative Neural Networks”, AIChE Journal, February 1991, Vol. 37, No. 2, pp. 233-243 to Kramer (“Kramer”, Published on Feb. 1991).
In regards to claim 1, Kim teaches the following:
1. A computer-implemented method for underwriting and pricing of loans for rideshare vehicles, the method comprising:
collecting information from third parties about the rideshare driver;
(See Kim, para. [0124]: “It is contemplated that, in some embodiments, an artificial intelligence model can be configured to accurately perform one or more of the steps comprising automatically underwriting loans, as described herein. For example, in some embodiments, automatic underwriting module 3570 may further include an artificial intelligence (AI) module that is configured to train one or more AI (e.g., machine learning or other suitable AI) models to perform one or more steps of the invention. In some embodiments, the AI module can be implemented using any of various techniques including, without any limitation, case-based reasoning, rule-based systems, fuzzy models, genetic algorithms, cellular automata, multi-agent systems, swarm intelligence, reinforcement learnings, artificial neural networks, hybrid systems, and the like. In one embodiment, the AI module may employ an artificial neural network to obtain and process fragmented pieces of information, as mentioned above. In some embodiments, the AI module may be trained on dataset(s) representing user account histories and/or previous readjustments, determinations, qualifying offers, weights, metrics, thresholds, qualifying scores, or any other suitable or relevant training data. In some embodiments, the AI module may be trained on the specific user's account(s) and/or credit history.”)
utilizing a base layer underwriting model to make a default risk prediction based on the information collected from the rideshare driver and the third parties;
(See Kim, para. [0124]: “It is contemplated that, in some embodiments, an artificial intelligence model can be configured to accurately perform one or more of the steps comprising automatically underwriting loans, as described herein. For example, in some embodiments, automatic underwriting module 3570 may further include an artificial intelligence (AI) module that is configured to train one or more AI (e.g., machine learning or other suitable AI) models to perform one or more steps of the invention. In some embodiments, the AI module can be implemented using any of various techniques including, without any limitation, case-based reasoning, rule-based systems, fuzzy models, genetic algorithms, cellular automata, multi-agent systems, swarm intelligence, reinforcement learnings, artificial neural networks, hybrid systems, and the like. In one embodiment, the AI module may employ an artificial neural network to obtain and process fragmented pieces of information, as mentioned above. In some embodiments, the AI module may be trained on dataset(s) representing user account histories and/or previous readjustments, determinations, qualifying offers, weights, metrics, thresholds, qualifying scores, or any other suitable or relevant training data. In some embodiments, the AI module may be trained on the specific user's account(s) and/or credit history.”)
refining the default risk prediction from the base layer underwriting model with available alternative data about the rideshare driver including the rental history and any specific business conditions;
(See Kim, para. [0125]: “In some embodiments, the AI module may be configured to self-correct and improve based on a reward feedback circuit for modifying the one or more assigned weights and/or unique thresholds. For example, upon a readjustment of the weights of metrics, the readjustment values can be measured as leading to a desired outcome or undesired outcome according to one or more criteria, then fed back into the AI module as training data. As the AI module further trains on the readjustments it makes and whether those readjustments are desirable or undesirable, the AI module self-corrects and improves as it continues to make readjustments. A wide variety of potential AI module can be employed for such a self-correcting technique. In some embodiments, one or more AI models comprising the AI module may be used in a similar manner for other determinations, computations, and/or predictions within the system. In some embodiments, one or more AI models comprising the AI module may be configured to automatically generate any one or more of metrics, weights, thresholds, tiers, qualifying scores for tiers, qualifying offers, reports, or any other suitable aspect of the systems described herein, without limitation. In some embodiments, such AI models comprising the AI module may be configured to similarly self-correct and improve this generation based on feedback reward circuits.”)
approving or denying a loan based on the risk prediction;
(See Kim, para. [0004]: “Lending, particularly originating loans such as personal loans, requires many fragmented, often manual processes of both borrowers and lenders. For a borrower, such processes include filling out a loan application and providing information in support of the loan application, the supporting information including, for example, employment, income, and liability information. For a lender, such processes include processing the borrower's loan application and verifying the supporting information, underwriting a potential loan and performing a detailed risk assessment in view of the supporting information, and, ultimately, upon approval from underwriting, funding the loan. Moreover, such processes are highly specific to loan type. This obviates any financial benefit from economies of scale that could otherwise be passed onto borrowers and lenders alike if such processes were more tightly integrated. Accordingly, there is a need for a more highly automated, more tightly integrated lending platform that facilitates lending for at least unsecured loan types such as personal loans.”)
However, under a conservative interpretation of Kim, it could be argued that Kim does not explicitly teach the following features, which are taught by Masson:
providing a loan pre-qualification application to a rideshare driver;
(See Masson, para. [0028]: “The Customer Data Input 202, the Business Data and Return Target 210, and the Economic Data 216, amongst other things, serves as a repository for storing data processed, received, and generated by one or more of the modules. The data may also include a data repository and other data. The system 102 may be accessed by the user device 106 registered with the system 102. The user device 106 may belong to an entity that may offer lending products to its customers (e.g. Financial Institution, Bank, Credit Union, Fintech, DeFi), a credit rating agency, a regulatory institution, an asset management business, a private equity business, a wealth management business, or an equity research business. Further, each of the aforementioned modules is explained in subsequent paragraphs of the specification.”)
(See Masson, para. [0029]: “Now referring to FIG. 2 , the Customer Data Input 202 may provide customer data to the system 102, such as but not limited to customer characteristics, loan application data, lending performance data, transactional data, and other data. The data may be sourced from the financial institution itself, third-party vendors, or data providers.”)
collecting credit and rental history information from the rideshare driver;
…
(See Masson, para. [0029]: “Now referring to FIG. 2 , the Customer Data Input 202 may provide customer data to the system 102, such as but not limited to customer characteristics, loan application data, lending performance data, transactional data, and other data. The data may be sourced from the financial institution itself, third-party vendors, or data providers.”)
(See Masson, para. [0036]: “The Customer Data Input 202 may be imported into the system 102 from a simple data file manually uploaded by the user from the interfaces 230 (e.g., Financial Institution side 704 / vendor side 718) or through API connections (e.g., Financial Institution side 702 / vendor side 716). The type of data imported may contain customer characteristics 706 / 720 (e.g. name, address, income, financial institution relationship), loan application data 708 / 722 (e.g. product, amount financed, term, pricing, application credit score, length of credit history), lending performance data 710 / 724 (e.g. payment status, balance, monitoring credit score), transactional data 712 / 726 (e.g. retail store detailed transaction spend, cash advanced transactions, debt payment transactions) and any other data 714 / 728 that may be used for lending strategies (e.g. alternative data, utility and phone bills and payments, court decisions, business reviews, social media presence and profiles).”)
optimizing the price of a loan and win probability at a desired yield by employing a price optimization and win probability model based on the refined outputs of the underwriting model,
(See Masson, para. [0034]: “At step 508, the processor 228 may create new datasets from the Customer Data Input 202. These new datasets may be the result of the application of the rules 318 defined in the credit strategy 325, pricing strategy 328 (e.g., Risk-based pricing), amount strategy 330, and tenure strategy 332 for all automated lending strategies 316 of the automated strategy set 506 onto the Customer Data Input 202. This process may apply the rules 318 at the product and vintage level to account for any product or function specific strategies (e.g., marketing/acquisition, underwriting rules for the vintages less than 12 months, account management rules for the vintages older than 12 months). Any conflicting rule during this process may result in selecting the least favorable value or decision for the new data (e.g., loan decline decision, highest interest rate).”)
wherein the price optimization and win probability model comprises
…
determining the optimal price and optimal win probability from the final output constrained to follow the sigmoid relationship; and
utilizing a loan terms model to set loan terms that are market competitive.
(See Masson, para. [0004]: “The forecasting and stress testing module is configured to execute in tandem with an artificial intelligence / machine learning module to process and analyze the new datasets using the new strategies to automatically identify an optimized regression model to forecast business revenue or any other lending Key Process Indicators (KPIs) over a specific time window based on several new strategies, economic and business forecast and stress test scenarios. The business return tracking module is configured to process and deliver a detailed comparison between the user target business revenue and the optimized business revenue.”)
(See Masson, para. [0005]: “The program includes programmed instructions for processing and analyzing the data input to automatically identify the best lending strategy in terms of business revenue across all specific industry knowledge characteristics of lending dimensions and lending functions.”)
(See Masson, para. [0032]: “The Artificial Intelligence / Machine Learning module 326 may create lending strategies 316 based on several lending dimensions, which may include lending rules 318, vendor model evaluation 320, and internal model builder 322. These steps may provide the critical building blocks for each lending strategy, such as but not limited to Credit Strategy 325 to define the Credit Risk and Volume 334 components which may define the credit eligibility criteria for a customer to a program or a lending product (i.e. approval/decline rules), Pricing Strategy to define the Interest Rate and Fees 336 components, Amount Strategy 330 to define Line and Loan Amounts 338, Tenure Strategy to define the Life of a Loan 340 and loan term components.”)
It would have been obvious to a person having ordinary skill in the art (PHOSITA), before the effective filing date of the claimed invention, to include in the system and methods for personal loan-lending using artificial intelligence, as taught by Kim, in combination with the system and method of artificial intelligence based lending strategies, as taught by Masson, because both are in the same art of using artificial intelligence for making loans, and Masson discloses that such methods “optimize the lending business revenue” (See Masson’s Abstract).
However, under a conservative interpretation of Kim in view of Masson, it could be argued that Kim in view of Masson does not explicitly teach the following specific features, which are taught by a combination of Kramer and O’Shea:
wherein the price optimization and win probability model comprises
a dense feedforward neural network with a linear bypass; and
a final layer that translates dense nodes in hidden layers in the dense feedforward neural network into parameters for a sigmoid function and feeding hold-out variables into the sigmoid function to generate a final output constrained to follow the sigmoid relationship;
(See Kramer, page 235, col.1, para. 1: “Note that to achieve the universal fitting property, exactly one layer of sigmoidal nodes and two layers of weighted connections are required. In practice, sigmoidal nonlinearities are often included in the nodes of the output layer so that the network produces outputs in a fixed, finite range. Also, the sigmoid function can be scaled multiplicatively or translated without affecting the generality of the network. Since we frequently deal with mean-centered data sets, a sigmoid function scaled into the range (- 1, 1) was used in this work.”)
The Examiner interprets that Kramer’s “output layer” reads upon the claimed “final layer”,
(See O’Shea, para. [0048]: “In general, the machine-learning network 202 may include one or more collections of multiplications, divisions, and summations of inputs and intermediate values, optionally followed by non-linearities (such as rectified linear units, sigmoid function, or otherwise) or other operations (e.g. normalization), which may be arranged in a feed-forward manner, including optional bypass or residual connections or in a manner with feedback and in-layer connections (e.g., a recurrent or quasi-recurrent network). Parameters and weight values in the network may be used for a single multiplication, as in a fully connected neural network (DNN), or they may be “tied” or replicated across multiple locations within the network to form one or more receptive fields, such as in a convolutional neural network, a dilated convolutional neural network, a residual network unit, or similar. A collection of one or more of these layers may constitute the machine-learning network 202. The specific structure for the networks may be explicitly specified at design time, or it may be selected from a plurality of possible architecture candidates to ascertain the best performing candidate.”)
It would have been obvious to a person having ordinary skill in the art (PHOSITA), before the effective filing date of the claimed invention, to include in the system and methods for personal loan-lending using artificial intelligence, as taught by Kim, in combination with the system and method of artificial intelligence based lending strategies, as taught by Masson, because both are in the same art of using artificial intelligence for making loans, and Masson further discloses that such methods “optimize the lending business revenue” (See Masson’s Abstract), and further in view of Kramer and O’Shea, because Kim and Masson teach the use of artificial intelligence in loan underwriting, while Kramer and O’Shea teach specific implementations of artificial intelligence .
In regards to claim 2, Kim teaches:
2. The method of claim 1, further comprising
tracking parameters about the loan including payment schedule and payments made;
enabling the rideshare driver ability to make payments; and
updating the loan information regarding payments.
(See Kim, para. [0017]: “In some embodiments, the borrower-related information from the banking information section transferred to the database server and stored in the database on the storage device of the at least one server host of the one or more server hosts is later used by the personal loan-servicing system for automatically setting up monthly Automated Clearing House (“ACH”) payments on personal loans in accordance with terms of the personal loans, the terms ranging from 3 to 5 years.”)
(See Kim, para. [0029]: “In some embodiments, the number of steps further include automatically setting up monthly ACH payments by way of the personal loan-originating system on personal loans in accordance with terms of the personal loans. The borrower-related information from the banking information section transferred to the database server and stored in the database on the storage device of the at least one server host of the one or more server hosts is used for automatically setting up the monthly ACH payments.”)
In regards to claim 5, Kim teaches:
The method of claim 1, wherein the collecting credit and rental history information from the rideshare driver includes the rideshare driver granting bank account access to a third-party financial services company.
(See Kim, para. [0072]: “The borrower-related information from the banking information section of the digital application transferred to the database server and stored in the one or more databases 3540 on the storage device of the at least one server host of the one or more server hosts can be used by the personal loan-servicing system 2500 for automatically setting up monthly ACH payments on personal loans in accordance with terms of the personal loans, which terms range from 3 to 5 years.”)
In regards to claim 6,
6. (New) The method of claim 1, wherein the base underwriting model receives updated borrower-specific data from the telematics device and retrains the neural-network parameters based on the updated data to refine the price optimization output.
(See Kim, para. [0096]: “When used in a LAN networking environment, the network host 800 is connected to the LAN 871 through a network interface or adapter 870, which can be, for example, a Bluetooth® or Wi-Fi adapter. When used in a WAN networking environment (e.g., Internet), the network host 800 can include some means for establishing communications over the WAN 873. With respect to telecommunication technologies, for example, a radio interface, which can be internal or external, can be connected to the system bus 821 via the network interface 870, or another appropriate mechanism. In a networked environment, other software depicted relative to the network host 800, or portions thereof, can be stored in the remote memory storage device. By way of example, and not limitation, FIG. 8 illustrates remote application programs 885 as residing on the network host 880. It will be appreciated that the network connections shown are examples and other means of establishing a communications link between the network hosts can be used.”)
In regards to claim 7,
7. (New) The method of claim 1, further comprising transmitting a vehicle-status signal from the telematics device to a server executing the underwriting model,
(See Kim, para. [0096]: “When used in a LAN networking environment, the network host 800 is connected to the LAN 871 through a network interface or adapter 870, which can be, for example, a Bluetooth® or Wi-Fi adapter. When used in a WAN networking environment (e.g., Internet), the network host 800 can include some means for establishing communications over the WAN 873. With respect to telecommunication technologies, for example, a radio interface, which can be internal or external, can be connected to the system bus 821 via the network interface 870, or another appropriate mechanism. In a networked environment, other software depicted relative to the network host 800, or portions thereof, can be stored in the remote memory storage device. By way of example, and not limitation, FIG. 8 illustrates remote application programs 885 as residing on the network host 880. It will be appreciated that the network connections shown are examples and other means of establishing a communications link between the network hosts can be used.”)
wherein the signal is processed to modify model weighting of payment-history features.
(See Kim, para. [0051]: “It is contemplated that, in some embodiments, the lending platform 1100 may include an artificial intelligence (AI) model that can be configured to accurately perform one or more of the steps comprising automatically underwriting loans, as described herein. For example, the server host 5020 (see FIG. 5 ) can include an AI module that may be configured to train one or more AI (e.g., machine learning or other suitable AI) modules to perform one or more steps of the invention. In some embodiments, a neural network may be utilized. In some embodiments, the AI modules may be trained on dataset(s) representing user account histories and/or previous readjustments, determinations, qualifying offers, weights, metrics, thresholds, qualifying scores, or any other suitable or relevant training data.”)
In regards to claim 8,
8. (New) The method of claim 1, wherein the underwriting model comprises a base model trained using gradient-boosting techniques
(See O’Shea, para. [0077]: “The update process 516 may utilize various techniques to determine a suitable update of the machine-learning network 502, such as an optimization method including evolution, gradient descent, stochastic gradient descent, or other solution techniques. In some implementations, the update process 516 may include user preferences or application specifications.”)
and a refinement process employing Monte-Carlo sampling to generate feature distributions used as inputs to the dense neural network.
(See O’Shea, para. [0077]: “In some implementations, one or more RF signals from the training dataset 518 may be augmented to include additional effects, for example to model real-world phenomena or variations in signaling (e.g., hardware imperfections, wireless propagation, or other variations). For example, the particular RF signal 504 may be stored and labeled (e.g., with stored label information 526) in the training dataset 518. During training, when the RF signal 504 is processed by the machine-learning network 502, one or more additional effects may be introduced in the RF signal 504 to broaden the types or number of examples of signals that are modelled by the RF signal 504. For example, the system may introduce random phase/frequency offsets, time offsets, time dilations, delay spreads, fading effects, distortion effects, interference effects, spatial propagation effects, dispersion effects, differing contents, non-linear effects, noise, and/or other signal effects in the RF signal. Such effects may be implemented, for example, as regularization layers in the machine-learning network 502 or as other augmentation effects while training or while pre-processing the input data prior to training. As such, the training may be made more robust (e.g., to generalize well) to identify not only the particular RF signal 504, but also to identify a range of RF signals that correspond to the RF signal 504 having been affected or perturbed by real-world variability (e.g., through a plurality of different propagation modes or differing channel state information).”)
Claim 3 is rejected under 35 U.S.C. 103 as being unpatentable over Kim in view of Masson, Kramer, and O’Shea as applied to independent claim 1 above, and further in view of “Miss a Payment? Good Luck Moving That Car” by Michael Corkery et al. (“Corkery”. Published Sept. 24, 2014).
In regards to claim 3, under a conservative interpretation of Kim in view of Masson, Kramer, and O’Shea, it could be argued that Kim in view of Masson, Kramer, and O’Shea do not explicitly teach the features below, which are taught by Corkery:
3. The method of claim 2, further comprising
installing a telematics device on the vehicle financed with the loan,
(See Corkery, page 2: “But before they can drive off the lot, many subprime borrowers like Ms. Bolender must have their car outfitted with a so-called starter interrupt device, which allows lenders to remotely disable the ignition. Using the GPS technology on the devices, the lenders can also track the cars’ location and movements.”)
wherein the telematics device is capable of tracking location of the vehicle, communicating with the rideshare driver and disabling ignition of the vehicle in event of loan default.
(See Corkery, page 2: “But before they can drive off the lot, many subprime borrowers like Ms. Bolender must have their car outfitted with a so-called starter interrupt device, which allows lenders to remotely disable the ignition. Using the GPS technology on the devices, the lenders can also track the cars’ location and movements.”)
It would have been obvious to a person having ordinary skill in the art (PHOSITA), before the effective filing date of the claimed invention, to include the telematics device taught by Corkery, with the system and methods for personal loan-lending using artificial intelligence, as taught by Kim, in combination with the system and method of artificial intelligence based lending strategies, as taught by Masson, because both are in the same art of using artificial intelligence for making loans, and Masson discloses that such methods “optimize the lending business revenue” (See Masson’s Abstract), and because as Corkery teaches that “The devices, which have been installed in about two million vehicles, are helping feed the subprime boom by enabling more high-risk borrowers to get loans”. (See Corkery , page 2).
The Examiner interprets that the devices enable easier repossession of the vehicle by the lender, in cases where the borrower defaults on the loan. As further taught by Corkery (see pages 2 and 3):
“But there is a big catch. By simply clicking a mouse or tapping a smartphone, lenders retain the ultimate control. Borrowers must stay current with their payments, or lose access to their vehicle”, and “And without them, they say, millions of Americans might not qualify for a car loan at all.”
Claims 4 is rejected under 35 U.S.C. 103 as being unpatentable over Kim in view of Masson, Kramer, and O’Shea as applied to independent claim 1 above, and further in view of Official Notice.
In regards to claim 4, while Kim teaches “automatically setting up monthly Automated Clearing House (“ACH”) payments on personal loans” in Kim’s para. [0017] and [0029], under a conservative interpretation of Kim in view of Masson, Kramer, and O’Shea, it could be argued that these references do not explicitly teach the features below, which are rejected on the basis of Official Notice:
4. The method of claim 1, wherein the loan terms include weekly payments.
It would have been obvious to a person having ordinary skill in the art (PHOSITA), before the effective filing date of the claimed invention, to in addition to enabling “weekly payments” in addition to “automatically setting up monthly Automated Clearing House (“ACH”) payments on personal loans”, as taught by Kim in Kim’s para. [0017] and [0029], because Official Notice is given that bi-weekly and weekly loan payments are well-known obvious variations to monthly payments.
Response to Amendments
Re: Drawings
The objections to the drawings are withdrawn, in response to the submission of Replacement Figure 4.
Re: Election/Restrictions
Newly added claims 9-15 have been issued a restriction by original presentation, due to these new claims being directed to inventions that are independent or distinct from the invention originally claimed.
Re: Claim Rejections - 35 USC § 112(b)
New 35 USC § 112(b) rejections have been applied to newly added dependent claims 6 and 7. The new rejections are necessitated by Applicant’s amendments to the claims.
Re: Claim Rejections - 35 USC § 101
The 35 USC § 101 rejection has been amended. The new rejection is necessitated by the DesJardins decision.
The rejection has been amended to recite that in an alternative interpretation, “the machine learning model (i.e. the neural network) is an additional element that is merely being run on the general purpose computer, and applied to the abstract idea (loans terms, risk of default).”
Furthermore, the Examiner finds that Applicant’s arguments in pages 11-15 of the response filed on Dec. 8, 2026 are not persuasive.
More specifically, the Examiner continues to hold that independent claim 1 recites “Certain Methods of Organizing Human Activity", specifically “Commercial or Legal Interactions (Including Agreements in the form of Contracts; Legal Obligations; Advertising, Marketing, or Sales Activities or Behaviors; Business Relations)”, as discussed in MPEP §2106(a)(2) Parts (I) and (II), and in the 2019 Revised Patent Subject Matter Eligibility Guidance, and also recite “Mathematical Concepts", specifically “Mathematical Relationships”, “Mathematical Formulas or Equations”, and “Mathematical Calculations”, as discussed in MPEP §2106.04(a)(2) Part (IV), and in the 2019 Revised Patent Subject Matter Eligibility Guidance.
More specifically, the Examiner interprets that the claimed “linear bypass connections” and “sigmoid-constraint layer” in independent claim 1 are “Mathematical Concepts", specifically “Mathematical Relationships”, “Mathematical Formulas or Equations”, because claim 1 in its entirety is directed to “Commercial or Legal Interactions” and “Mathematical Relationships”.
Re: Claim Rejections - 35 USC § 103
The 35 USC § 103 rejection has been amended. The new rejections are necessitated by the newly added claims and by the newly discovered prior art references.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
See US-2022/0414685-A1 to Lamba et al:
[0095] Dense Neural Network: The proposed ensemble method jointly learns multiple identical neural networks, one model per time window. The individual components in the ensemble follow the multilayer perceptron approach, which implements a feedforward artificial neural network (ANN) architecture consisting of multiple layers of perceptrons or dense nodes. Each neural network's input layer contains (N=number of events in the considered sentiment category) nodes indicating the term frequencies of negative events for the corresponding window. For example, if negative events are considered, the input layers may have 928 nodes. This may be followed by a dense hidden layer with 60 neurons. The hidden layer may be activated by using a Rectified Linear Unit (ReLu) that is responsible for transforming the summed weight of a layer's input into its output, thus resulting in an embedding for each time window. The outputs of the hidden layers from all networks are then stacked together and passed as an input to downstream prediction tasks. This can be done by adding a final output layer to the ensemble that has two nodes with a Sigmoid activating function. The input size of the final output layer is 180 (i.e., 60 neurons from each component's hidden layer).
See the Abstract of US 2025/0173787 A1 to Kim et al.:
This disclosure relates to a lending system comprising a processor, a memory, and at least one network interface controller configured to enable data exchange with external systems. The memory includes lending management logic configured to execute a personal loan-lending system wherein one or more loans are generated by collecting and preprocessing borrower data from multiple sources, conducting risk assessment and scoring, generating one or more scores based on the conducted risk assessment, and approving loans based on the generated scores. The system leverages machine learning processes to refine scoring accuracy and dynamically adapt to borrower profiles and external conditions. The system may integrate with third-party data sources for enhanced verification and incorporate compliance logic to ensure adherence to regulatory standards. By employing automated, data-driven processes, the system improves efficiency, accuracy, and security in loan origination, servicing, and compliance while enabling dynamic risk management and tailored loan product offerings.
Applicants are invited to contact the Office to schedule an in-person interview to discuss and resolve the issues set forth in this Office Action. Although an interview is not required, the Office believes that an interview can be of use to resolve any issues related to a patent application in an efficient and prompt manner.
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.
Any inquiry concerning this communication or earlier communications should be directed to Examiner Ayal Sharon, whose telephone number is (571) 272-5614, and fax number is (571) 273-1794. The Examiner can normally be reached from Monday to Friday between 9 AM and 6 PM. If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, SPE Christine Behncke can be reached at (571) 272-8103 or at christine.behncke@uspto.gov. The fax 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.
Sincerely,
/Ayal I. Sharon/
Examiner, Art Unit 3695
January 16, 2026