Systems and methods for prioritizing fraud cases using artificial intelligence

1 . A method comprising:

receiving, by a processing circuit, sets of fraud cases based on a trigger event, each fraud case relating to one or more transactions and comprising transaction data comprising authentication information indicating a type of authentication used in the one or more transactions and an initial priority score, wherein the trigger event indicates the initial priority score exceeds a predetermined threshold;

generating, by the processing circuit, a plurality of category queues based on ranking each case of the sets of fraud cases, each category queue corresponding to a different type of fraudulent activity, wherein each category queue is sorted based on the initial priority score of each case of the sets of fraud cases;

determining, by the processing circuit, an updated priority score for each fraud case based on the transaction data and case prioritization data, the case prioritization data comprising rules developed using a machine learning model, the machine learning model having been trained using training data comprising test data and validation data associated with previous fraud cases and historical transaction data,

wherein the updated priority score is determined by applying an algorithm corresponding to the rules,

wherein one or more of the rules are based on the type of authentication used in the one or more transactions; and

wherein a highest priority fraud case is dynamically determined in real-time based on the updated priority score by ranking each fraud case within each category queue;

re-sorting, by the processing circuit, the plurality of category queues based on the updated priority score;

assembling, by the processing circuit, a first group of the fraud cases associated with a first risk level of a first queue in a central case database based on the re-sorted plurality of category queues, wherein the fraud cases are stored in the central case database to facilitate faster identification of fraud cases and retrieval of corresponding data, wherein each fraud case is sorted within the central case database based on the updated priority score, and wherein data corresponding to the highest priority fraud case is identified and moved to a location within the central case database to facilitate faster identification and assignment of the highest priority fraud case; and

performing, by the processing circuit, an action on a second group of the fraud cases associated with a second risk level of a second queue based on the re-sorted plurality of category queues;

wherein the machine learning model is retrained based on the transaction data and the case prioritization data is restructured based on the retrained machine learning model, and the machine learning model comprises a classification model configured to associate each fraud case with the first group of the fraud cases or the second group of the fraud cases.

2 . The method of claim 1 , wherein the sets of fraud cases are received from a fraud identification system, and wherein the fraud identification system assigns the initial priority score.

3 . The method of claim 1 , wherein the updated priority score is higher than the initial priority score, lower than the initial priority score, or the same as the initial priority score.

4 . The method of claim 1 , wherein each of the sets of fraud cases corresponds to a case category, the case category comprising at least one of a credit card transaction category, a debit card transaction category, an online transaction category, or an identity theft category.

5 . The method of claim 1 , further comprising:

receiving, by the processing circuit from a fraud agent computing terminal, a request to return the highest priority fraud case;

determining, by the processing circuit, the highest priority fraud case by identifying a fraud case in the first group of the fraud cases having a highest updated priority score; and

transmitting, by the processing circuit, the highest priority fraud case to the fraud agent computing terminal.

6 . The method of claim 1 , further comprising:

receiving, by the processing circuit, the sets of fraud cases in response to receiving the trigger event, wherein the trigger event relates to a reoccurring time period or a potentially fraudulent transaction.

7 . The method of claim 1 , further comprising:

storing, by the processing circuit, the sets of fraud cases, the updated priority score, and the case prioritization data in the central case database.

8 . The method of claim 1 , wherein the machine learning model comprises at least one of a supervised learning model, an unsupervised learning model, or a reinforcement learning model.

9 . A system comprising:

a processing circuit configured to:

receive sets of fraud cases based on a trigger event, each fraud case relating to one or more transactions and comprising transaction data comprising authentication information indicating a type of authentication used in the one or more transactions and an initial priority score, wherein the trigger event indicates the initial priority score exceeds a predetermined threshold;

generate a plurality of category queues based on ranking each case of the sets of fraud cases, each category queue corresponding to a different type of fraudulent activity, wherein each category queue is sorted based on the initial priority score of each case of the sets of fraud cases;

determine an updated priority score for each fraud case based on the transaction data and case prioritization data, the case prioritization data comprising rules developed using a machine learning model, the machine learning model having been trained using training data comprising test data and validation data associated with previous fraud cases and historical transaction data, wherein the updated priority score is determined by applying an algorithm corresponding to the rules,

wherein one or more of the rules are based on the type of authentication used in the one or more transactions; and

wherein a highest priority fraud case is dynamically determined in real-time based on the updated priority score by ranking each fraud case within each category queue;

re-sort the plurality of category queues based on the updated priority score;

assemble a first group of the fraud cases associated with a first risk level of a first queue in a central case database based on the re-sorted plurality of category queues, wherein the fraud cases are stored in the central case database to facilitate faster identification of fraud cases and retrieval of corresponding data, wherein each fraud case is sorted within the central case database based on the updated priority score, and wherein data corresponding to the highest priority fraud case is identified and moved to a location within the central case database to facilitate faster identification and assignment of the highest priority fraud case; and

perform an action on a second group of the fraud cases associated with a second risk level of a second queue based on the re-sorted plurality of category queues;

wherein the machine learning model is retrained based on the transaction data and the case prioritization data is restructured based on the retrained machine learning model, and the machine learning model comprises a classification model configured to associate each fraud case with the first group of the fraud cases or the second group of the fraud cases.

10 . The system of claim 9 , wherein the sets of fraud cases are received from a fraud identification system, and wherein the fraud identification system assigns the initial priority score.

11 . The system of claim 9 , wherein the updated priority score is higher than the initial priority score, lower than the initial priority score, or the same as the initial priority score.

12 . The system of claim 9 , wherein each of the sets of fraud cases corresponds to a case category, the case category comprising at least one of a credit card transaction category, a debit card transaction category, an online transaction category, or an identity theft category.

13 . The system of claim 9 , wherein the processing circuit is further configured to:

receive, from a fraud agent computing terminal, a request to return the highest priority fraud case;

determine the highest priority fraud case by identifying a fraud case in the first group of the fraud cases having a highest updated priority score; and

transmit the highest priority fraud case to the fraud agent computing terminal.

14 . The system of claim 9 , wherein the processing circuit is further configured to:

receive the sets of fraud cases in response to receiving the trigger event, wherein the trigger event relates to a reoccurring time period or a potentially fraudulent transaction.

15 . The system of claim 9 , wherein the processing circuit is further configured to:

store the sets of fraud cases, the updated priority score, and the case prioritization data in the central case database.

16 . The system of claim 9 , wherein the machine learning model comprises at least one of a supervised learning model, an unsupervised learning model, or a reinforcement learning model.

17 . One or more non-transitory computer-readable storage media having instructions stored thereon that, when executed by one or more processors, cause the one or more processors to:

receive sets of fraud cases based on a trigger event, each fraud case relating to one or more transactions and comprising transaction data comprising authentication information indicating a type of authentication used in the one or more transactions and an initial priority score, wherein the trigger event indicates the initial priority score exceeds a predetermined threshold;

generate a plurality of category queues based on ranking each case of the sets of fraud cases, each category queue corresponding to a different type of fraudulent activity, wherein each category queue is sorted based on the initial priority score of each case of the sets of fraud cases;

determine an updated priority score for each fraud case based on the transaction data and case prioritization data, the case prioritization data comprising rules developed using a machine learning model, the machine learning model having been trained using training data comprising test data and validation data associated with previous fraud cases and historical transaction data,

wherein the updated priority score is determined by applying an algorithm corresponding to the rules,

wherein one or more of the rules are based on the type of authentication used in the one or more transactions; and

wherein a highest priority fraud case is dynamically determined in real-time based on the updated priority score by ranking each fraud case within each category queue;

re-sort the plurality of category queues based on the updated priority score;

assemble a first group of the fraud cases associated with a first risk level of a first queue in a central case database based on the re-sorted plurality of category queues, wherein the fraud cases are stored in the central case database to facilitate faster identification of fraud cases and retrieval of corresponding data, wherein each fraud case is sorted within the central case database based on the updated priority score, and wherein data corresponding to the highest priority fraud case is identified and moved to a location within the central case database to facilitate faster identification and assignment of the highest priority fraud case; and

perform an action on a second group of the fraud cases associated with a second risk level of a second queue based on the re-sorted plurality of category queues;

wherein the machine learning model is retrained based on the transaction data and the case prioritization data is restructured based on the retrained machine learning model, and the machine learning model comprises a classification model configured to associate each fraud case with the first group of the fraud cases or the second group of the fraud cases.