A method and system identify assistance offerings that are likely to be relevant to users of a data management system. The method and system utilize a multivariate random forest regression machine learning process to train an assistance offerings recommendation model to recommend relevant assistance offerings to users of the data management system. The multivariate random forest regression machine learning process replaces zero values in the training set data with negative numbers to increase the accuracy of the machine learning process.

Described herein are systems and techniques for providing automatic distribution of transaction-related benefits to an account used to conduct a transaction. In certain embodiments, the system receives transaction details from a transaction processing network as transactions are conducted. The system determines an appropriate benefit provider for a transaction conducted from an account and provides a subset of transaction details to that benefit provider. The benefit provider compares the subset of transaction details to a set of limited transaction details received from a resource provider to identify the transaction and determine a benefit amount, if any, for which the transaction qualifies. The benefit provider then provides the benefit amount to the system, which then causes the transaction processing network to credit the account the benefit amount.

A method optimizes questions to retain engagement. The method includes generating, using a machine learning model, a churn risk from user interaction data. The method includes selecting, when the churn risk satisfies a threshold, a field, from multiple fields, using multiple prediction confidences corresponding to multiple prediction values generated for the multiple fields. The method includes obtaining a prediction value for the field and obtaining a question, corresponding to the field, using the prediction value. The method includes presenting the question and receiving a user input in response to the question.

An authentication computing device stores a cardholder profile that is associated with a candidate cardholder and includes a cardholder identifier, a device identifier, payment account data, and trusted authentication data in a database system, receives an authentication request that is associated with a tax filing of the candidate cardholder and includes a filing identifier from a revenue computing device, detects the authentication request is associated with the candidate cardholder based upon the filing identifier and the cardholder profile, transmits an identity challenge requesting authentication data associated with the candidate cardholder to a user device associated with the device identifier, receives a challenge response including response authentication data from the user device, determines an authentication status associated with the authentication request based on a comparison of the response authentication data and the trusted authentication data, and transmits the authentication status to the revenue computing device.

A physical overlay instrument is described that is manufactured for use with a particular form, such as a specific version of a tax form. The overlay instrument comprises a sheet of material including transparent regions and opaque regions positioned such that when the overlay instrument is placed over the particular form filled out by a user, a first set of personal information of the user included on the form is visible through the transparent regions while as second set of personal information of the user included on the form is obscured by the opaque regions. In addition, the opaque regions include one or more prompts to gather additional information from the user based on the first set of the personal information of the user that is visible through the transparent regions. The overlay instrument is manufactured from a single sheet of material and may be one-sided or two-sided.

Systems and methods disclosed herein provide automatic expense categorization of transactions or expenditures based on a machine learning (ML) model trained using anonymized transactional data for expenditures that are stored in a public blockchain. The anonymized transactional data for the expenditure and the expense category, may be distributed throughout the blockchain network and recorded in the blockchain. In some implementations, an expenditure may be submitted to the blockchain network for expense categorization. The transactional data for the expenditure may be anonymized to remove confidential and personal identifying information (PII) before it is distributed throughout the blockchain network. Each participating node of the blockchain network may utilize the ML model to identify an expense category for the expenditure. The participating nodes may provide a consensus mechanism in order to arrive at a shared understanding of how to categorize the expenditure.

Systems and methods of taxation analysis for asset transactions are disclosed. A system may be configured to obtain data associated with a transaction to be executed, obtain a request to determine a tax implication of the transaction, generate a first tax return (with the transaction being excluded from use in generating the first tax return), and generate a second tax return based on the request (with the transaction being included for use in generating the second tax return). The system may also be configured to compare the first tax return and the second tax return, generate one or more results associated with the transaction based on the comparison, and provide the one or more results. The one or more results may indicate an impact on total taxes owed, deductions, loss of tax credits, or other impacts caused by the proposed transaction.

A method implements calibrated risk scoring and sampling. Features are extracted from a record. A risk score, associated with the record, is generated from the features using a machine learning model. The record is mapped to a risk bucket using the risk score. The risk bucket may include multiple risk bucket records. The record is selected from the risk bucket records with a sampling threshold corresponding to the risk bucket. A form prepopulated with values from the record is presenting to a client device.

A computer-implemented method comprising receiving a set of rules that define assigning tax codes for a first entity based on a plurality of parameters, and storing the set of rules in a first data repository in association with an entity record of the first entity; receiving invoice data that defines an invoice directed to the first entity; automatically determining a commodity type applicable to the invoice data by digitally cross-referencing line items in the invoice data representing goods or services to community transaction data stored in a second data repository, the community transaction data comprising a plurality of different line item data for different invoices of other entities different from and unrelated to the first entity; automatically assigning tax codes to the invoice data based on the commodity type and the set of rules; and causing to display the commodity type and the tax codes in a graphical user interface.

A system may include a processor that may detect one or more loan indicators present in banking data or credit data associated with a user. The processor may then retrieve additional data associated with the user in response to detecting the one or more loan indicators, such that the additional data may include data acquired from a home assistant device, a wearable device, a computing device, or any combination thereof. The processor may then determine a loan probability associated with the user based on the banking data, the credit data, and the additional data. The processor may then determine a pre-approval loan amount based on the banking data and the credit data in response to the loan probability exceeding a threshold and automatically send a notification indicative of a pre-approval loan amount to a computing device associated with the user.