Granted Patent B2

US 12,462,297 · App. 18/639,135 · Granted Nov 4, 2025

Inventor: Sebastian Pascal (Marlborough, MA)

G06Q40/02H04L67/306

Abstract

Claims (44)

1 . An apparatus for tracking fraudulent activity that will impact financial and economic integrity, the apparatus comprising:

a user database;

at least a processor; and

a memory communicatively connected to the processor, the memory containing instructions configuring the processor to:

receive one or more identification data;

receive, from the user database, a user profile associated with one or more identification data;

receive, from the user database, one or more local fraud risk factors including geofence information;

generate a fraud risk score as a function of the user profile and one or more local fraud risk factors utilizing a multilayer neural network score generation module which comprises:

receiving training data, wherein the training data correlates a plurality of identification data and one or more local fraud risk factor data to a plurality of score data;

training the multilayer neural network iteratively by filtering, sorting, and classifying training data using supervised or unsupervised classifiers; retraining the multilayer neural network based on updated fraud outcomes and feedback from previous scoring iterations, thereby progressively improving fraud-risk scoring accuracy;

securely identify an individual based on the fraud risk score and one or more identification data;

initiate one or more security parameters;

generate an alert as a function of the user profile whose content and severity are determined by the user profile risk tier; and

record, in an immutable hash-chained distributed ledger, the fraud risk score, the invoked counter-measure, and a time stamp to create a tamper-evident audit trail.

2 . The apparatus of claim 1 , wherein identification data comprises biographical data.

3 . The apparatus of claim 1 , wherein identification data comprises geographic data.

4 . The apparatus of claim 1 , wherein generation of the score based on the identification data instantiates a machine learning model.

5 . The apparatus of claim 1 , wherein generation of the score based on the identification data instantiates a neural network.

6 . The apparatus of claim 1 , wherein securely identifying the individual based on the score and one or more identification data instantiates a machine learning model.

7 . The apparatus of claim 1 , wherein securely identifying the individual based on the score and one or more identification data instantiates a neural network.

8 . The apparatus of claim 1 , wherein one or more security parameters are stored within the user profile.

9 . Apparatus of claim 1 , wherein generation of the alert based on one or more security parameters comprises the score, a breakdown of the score, and the individual's score history.

10 . The apparatus of claim 9 , wherein the generated alert updates the associated user profile, stored in the user database.

11 . A method for tracking fraudulent activity that will impact financial and economic integrity, the method comprising:

receiving one or more identification data;

receiving, from the user database, a user profile associated with one or more identification data;

receiving, from the user database, one or more local fraud risk factors, including geofence information;

generating a fraud risk score as a function of a user profile and one or more local fraud risk factors utilizing a multilayer neural network score generation module which comprises:

receiving training data, wherein the training data correlates a plurality of identification data and one or more local fraud risk factor data to a plurality of score data;

training, iteratively, the multilayer neural network using the training data, wherein training includes filtering, sorting, and classifying the training data using supervised or unsupervised classifiers, and includes retraining the multilayer neural network with feedback from previous iterations of the multilayer neural network; and

generating the fraud risk score using the trained multilayer neural network;

securely identifying an individual as a function of the fraud risk score and the identification data;

initiating one or more security parameters;

generating an alert whose content and severity correspond to a risk tier stored in the user profile; and

recording, in an immutable hash-chained distributed ledger, the fraud-risk score, the invoked security parameters, and a time stamp to create a tamper-evident audit trail.

12 . The method of claim 11 , wherein identification data comprises biographical data.

13 . The method of claim 11 , wherein identification data comprises geographic data.

14 . The method of claim 11 , wherein generation of the score based on the identification data instantiates a machine learning model.

15 . The method of claim 11 , wherein generation of the score based on the identification data instantiates a neural network.

16 . The method of claim 11 , wherein securely identifying the individual based on the score and one or more identification data instantiates a machine learning model.

17 . The method of claim 11 , wherein securely identifying the individual based on the score and one or more identification data instantiates a neural network.

18 . The method of claim 11 , wherein one or more security parameters are stored within the user profile.

19 . The method of claim 11 , wherein generation of the alert based on one or more security parameters comprises the score, a breakdown of the score, and the individual's score history.

20 . The method of claim 19 , wherein the generated alert updates the associated user profile, stored in the user database.

Assignments (2)

Continuity (2)

References Cited (26)