Framework for generating risk evaluation models

1. A device, comprising:

a non-transitory memory; and

one or more hardware processors coupled with the non-transitory memory and configured to read instructions from the non-transitory memory to cause the device to perform operations comprising:

obtaining, from a remote database system, a set of transaction records associated with a set of transactions;

retrieving, for each particular transaction record in the set of transaction records from the remote database system, additional transaction data associated with at least one other transaction record having a common attribute with the particular transaction record;

generating, based on the additional transaction data, an enriched set of transaction records;

serializing the enriched set of transaction records;

without further accessing the remote database system:

evaluating a plurality of features for performing risk assessment on transactions based on the serialized enriched set of transaction records, and wherein at least one of the plurality of features is derived from at least a portion of the additional transaction data;

selecting, from the plurality of features, a subset of features for performing risk assessment based on the evaluating;

configuring a machine learning model to evaluate transaction requests based on the selected subset of features;

deserializing the serialized enriched set of transaction records; and

training the machine learning model using the deserialized enriched set of transaction records; and

evaluating a transaction request using the trained machine learning model.

2. The device of claim 1 , wherein the operations further comprise:

receiving a selection of a first transaction type, from a plurality of transaction types, from a user; and

determining the set of transactions from a plurality of transactions based on the first transaction type.

3. The device of claim 1 , wherein the set of transaction records are stored in the remote database system according to a first data structure, wherein the operations further comprise:

storing the serialized enriched set of transaction records on the device according to a second data structure.

4. The device of claim 3 , wherein the machine learning model is configured to ingest transaction data according to the first data structure, and wherein the operations further comprise:

storing the deserialized enriched set of transaction records on the device according to the first data structure; and

validating the machine learning model based on the deserialized enriched set of transaction records.

5. The device of claim 4 , wherein the serialized enriched set of transaction records is deserialized in a local memory of the device without accessing the remote database system.

6. The device of claim 1 , wherein the evaluating the plurality of features comprises providing an interface for a user to manipulate transaction data in the enriched set of transaction records.

7. The device of claim 1 , wherein the evaluating the plurality of features comprises:

calculating, for each particular feature in the plurality of features, a value based on at least a portion of transaction data associated with a transaction record in the enriched set of transaction records; and

determining whether the value is indicative for evaluating a transaction associated with the transaction record.

8. The device of claim 1 , wherein the set of transaction records comprises a first transaction record associated with a first transaction, wherein the first transaction record comprises information associated with a first user who is a payor in the first transaction, and wherein the generating the enriched set of transaction records comprises retrieving, for the first transaction record, transaction data associated with a second transaction record associated with a second transaction in which the first user is also a payor.

9. The device of claim 1 , wherein the operations further comprise evaluating the set of transaction records using the subset of features.

10. The device of claim 9 , wherein the evaluating the set of transaction records using the subset of features comprises:

dividing the serialized enriched set of transaction records into a plurality of groups of transaction records;

transmitting the plurality of groups of transaction records to a plurality of servers, respectively; and

causing each server in the plurality of servers to evaluate the subset of features by running the machine learning model on the corresponding group of transaction records.

11. A method comprising:

obtaining, from a remote database system, a set of transaction records associated with a set of transactions;

retrieving, for each particular transaction record in the set of transaction records from the remote database system, additional transaction data associated with at least one other transaction record having a common attribute with the particular transaction record;

generating an enriched set of transaction records based on the additional transaction data;

serializing the enriched set of transaction records to generate serialized transaction records;

providing an application programming interface (API) for evaluating a plurality of features based on analyzing the serialized transaction records;

selecting, from the plurality of features, a subset of features based on the evaluating;

configuring a machine learning model to evaluate transaction requests based on the selected subset of features;

deserializing the serialized transaction records;

training the machine learning model using the deserialized transaction records; and

predicting, using the trained machine learning model, a risk of a transaction request.

12. The method of claim 11 , further comprising validating the machine learning model.

13. The method of claim 12 , wherein the validating the machine learning model comprises:

dividing the serialized transaction records into a plurality of groups of serialized transaction records, wherein each group of serialized transaction records includes transaction data and additional data associated with a subset of the set of transaction records;

transmitting the plurality of groups of serialized transaction records to a plurality of servers, respectively; and

causing each server in the plurality of servers to use the machine learning model to evaluate the corresponding subset of the set of transaction records based on the derived set of features.

14. The method of claim 11 , further comprising:

receiving a selection of a first transaction type, from a plurality of transaction types, from a user; and

determining the set of transactions from a plurality of transactions based on the first transaction type.

15. The method of claim 11 , further comprising:

receiving, via the API, a definition of a feature;

retrieving transaction data and additional data associated with a first transaction record from the serialized transaction records;

calculating a value corresponding to the feature based on at least a portion of the transaction data and the additional data associated with the first transaction record; and

presenting the value on a device.

16. The method of claim 11 , wherein the evaluating the plurality of features, the selecting the subset of features, and the configuring the machine learning model are performed without accessing the remote database system.

17. A non-transitory machine-readable medium having stored thereon machine-readable instructions executable to cause a machine to perform operations comprising:

obtaining, from a remote database system, a set of transaction records associated with a set of transactions;

retrieving, for each particular transaction record in the set of transaction records from the remote database system, additional transaction data associated with at least one other transaction record having a common attribute with the particular transaction record;

generating an enriched set of transaction records based on the additional transaction data;

serializing the enriched set of transaction records;

without further accessing the remote database system:

evaluating a plurality of features for risk assessment on transactions based on the serialized enriched set of transaction records, wherein at least one of the plurality of features is derived based on at least a portion of the additional transaction data;

selecting, from the plurality of features, a subset of features for performing risk assessment based on the evaluating;

configuring a machine learning model to evaluate transaction requests based on the subset of features;

deserializing the serialized enriched set of transaction records; and

training the machine learning model using the deserialized enriched set of transaction records; and

evaluating a transaction request using the trained machine learning model.

18. The non-transitory machine-readable medium of claim 17 , wherein the operations further comprise:

receiving a selection of a first transaction type, from a plurality of transaction types, from a user; and

determining the set of transactions from a plurality of transactions based on the first transaction type.

19. The non-transitory machine-readable medium of claim 17 , wherein the set of transaction records are stored in the remote database system according to a first data structure, wherein the operations further comprise:

storing the serialized enriched set of transaction records on a device external to the remote database system according to a second data structure.

20. The non-transitory machine-readable medium of claim 19 , wherein the first data structure is a flat data structure, and wherein the second data structure is a hierarchical data structure.