Abstract: Systems and methods are provided to determine consensus values for duplicate fields in a document or form.

Abstract: Certain aspects of the present disclosure provide techniques for placing targeted messages in communications within a software application using machine learning models. An example method generally includes retrieving, from a repository, a data set of targeted messages. For each respective targeted message in the data set of targeted messages, an effectiveness score for a party associated with the respective targeted message, a distance score between the party associated with the respective targeted message and a host party, and a match score between the party associated with the respective targeted message and the host party. Based on the effectiveness score, the distance score, and the match score for each respective targeted message in the data set of targeted messages, a message is selected to be included in one or more communications by the host party. The one or more communications including the selected message are generated and output for transmission.

Abstract: Certain aspects of the present disclosure provide techniques for adaptively reducing the bit size of features in a training data set used to train a machine learning model. An example method generally includes receiving a data set to be used in training a machine learning model and a definition of the machine learning model to be trained. A reduced number of bits to represent features in the data set is determined based on values of each feature in the data set and the definition of the machine learning model. A reduced bit-size data set is generated by reducing a bit size of each feature in the data set according to the reduced number of bits, and the reduced bit-size data set is encrypted using a homomorphic encryption scheme. A machine learning model is trained based on the encrypted reduced bit-size data set.

Abstract: Systems and methods for predicting a future status of an entity using a recurrent neural network are described. A system is configured to obtain a first time series data associated with a first entity. The system is also configured to predict, by a recurrent neural network, a future equity status of the first entity based on the time series data. The system is also configured to provide an indication of the future equity status of the first entity, with the future equity status of the first entity to be indicated to a user. An equity status may be a cash status of an entity, and predicting the cash status of the first entity can be based on time series of a plurality of features associated with the first entity provided to the recurrent neural network. The recurrent neural network may be a long short-term memory network.

Abstract: Certain aspects of the present disclosure provide techniques for generating a human readable summary explanation to a user for an outcome generated by a complex machine learning model. In one embodiment, a risk assessment service can receive a request from a user in which a risk model of the risk assessment service performs a specific task (e.g., determining the level of risk associated with the user). Once the risk model determines the risk associated with the user, in order to comply with regulations from a compliance system, the risk model can provide a user with an explanation as to the outcome for transparency purposes.

Abstract: Techniques are disclosed for managing workflows using block chains. An entire workflow may be stored as a finite state machine on a block chain. Each data block of the block chain may correspond to a state and related operation of a workflow. Each data block of the block chain may be used as an input to a subsequent state of the workflow. Thus, any input may be used to compute the workflow even if the input is not designed to be used by a workflow. Nested block chains, each corresponding to a different workflow or different part of a workflow may be used to increase performance of computing the workflows.

Abstract: Certain aspects of the present disclosure provide techniques for intelligent grouping of travel data for review through a user interface. In one example, a method for providing grouped travel data to a user interface of an application includes receiving travel data from an application running on a remote device; generating one or more travel data-based features from the travel data thereby creating featurized travel data; applying a pattern mining technique to the featurized travel data to detect a plurality of patterns in the featurized travel data; for each trip record in the featurized travel data: determining a plurality of trip record groups in which the trip record falls based on the plurality of patterns; and adding the trip record to a trip record group of the plurality of trip record groups according to a prioritization scheme; and transmitting the trip record group to the application running on the remote device.

Abstract: Certain embodiments of the present disclosure provide techniques for dynamically and reliably scaling a data processing pipeline in a computing environment. The method generally includes receiving a definition of a data pipeline to be instantiated on a set of resources in a computing environment. The data pipeline is converted into a plurality of steps, each step being defined as one or more workers. The one or more workers are instantiated. Each worker generally includes a user process and a processing coordinator to coordinate termination of the user process. Communications are orchestrated between one or more data sources and the one or more workers. The one or more workers are terminated by invoking a termination coordination process exposed by the user process and the processing coordinator associated with each worker of the one or more workers.

Abstract: A workflow construction system for constructing automation workflows that automate user specific processes. The workflow construction system may include a template library including workflow templates and pre-configured attributes. The workflow template can accelerate the design and construction of custom automation workflows. An orchestration layer included in the workflow construction system will also improve the performance of systems that execute the automation workflows by dynamically scaling the processing capacity, memory, and storage of servers and other systems hosting the model file instances of the automation workflows to ensure the available resources meet the demands of users completing processes using the automation workflows.

Abstract: A method and processing system for managing user access to one or more resources is disclosed. A central service may receive an access change request message regarding a user. The access change request message may include a user identifier, a user role, and an access action for the user. Example access actions may include adding or removing user access with respect to a resource. The central service may determine which resources are associated with the user role and transmit one or more event messages to the resources to implement the access actions. The resources may send acknowledgement messages to the central service to confirm that the access actions have been completed.

Abstract: This disclosure provides systems, methods and apparatuses for identifying fraudulent accounts associated with an electronic payment service. In some implementations, a computing device may retrieve a data set including a number of attributes for each of a multitude of accounts, and may construct a plurality of different graphs each based on a unique set of the attributes. Each graph may include a plurality of nodes linked together by a multitude of edges, where each node identifies a corresponding account and each edge indicates one or more of the corresponding attributes that are common to a pair of accounts. The computing device may determine a likelihood of each graph correctly identifying fraudulent accounts by analyzing groups of nodes connected to each other by corresponding groups of edges using historical account data, and may select the graph having the greatest determined likelihood to predict whether any of the accounts is fraudulent.

Abstract: Systems, methods and articles of manufacture for performing a comparison of tax results based on different sets of tax data, and generating an explanation as to why the tax results differ or do not differ due to the differences in the tax data. The system includes a computing device, a data store in communication with the computing device and a tax preparation software application executable by the computing device. The tax preparation software application has a tax calculation engine, a tax calculation graph, and a change analysis engine. The tax calculation engine is configured to perform a plurality of tax calculation operations based on the tax calculation graph. The change analysis engine is configured to determine whether tax results based on different tax data differ or do not differ. The system may also generate explanation(s) of the reasons that the tax results differ or do not differ due to the different tax data.

Abstract: A unified graph query system provides an abstraction layer that increases the interoperability of different graph technologies by exposing graphs stored in graph databases using a unified query language. The abstraction layer generates graph models for each of the available graph databases and extracts a graph component and other source data used to identify the source of the data requested by a query. The unified graph query system executes the query across the multiple graphs included in different graph databases by using the graph models to locate the graph component in each of the multiple graphs and extract the feature data associated with the graph component. The feature data is used to generate features that are used by a machine learning service to train machine learning models and is also used to make predictions in real time.

Abstract: A method including generating, by a state engine from data describing behaviors of users in an environment external to the state engine, an executable process. An agent executes the executable process by determining, from the data describing the behaviors of the users, a problem of at least some of the users, and selects, based on the problem, a chosen action to alter the problem. At a first time, a first electronic communication describing the chosen action to the at least some of the users is transmitted. Ongoing data describing ongoing behaviors of the users is monitored. A reward is generated based on the ongoing data to change a parameter of the agent. The parameter of the agent is changed to generate a modified agent. The modified agent executes the executable process to select a modified action. At a second time, a second electronic communication describing the modified action is transmitted.

Abstract: A content editor for generating content including root blocks and nested blocks is disclosed. The content editor can generate a deployment that includes the content. The content editor can generate user interface code configured to edit the content. The content editor can receive updates to the content and update the root blocks and nested blocks. The updated root blocks and nested blocks can be used to generate updated content for editing and/or can be deployed to end-users.

Abstract: 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.

Abstract: 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.

Abstract: A method of determining whether a user has a self-control problem. A computing system receives, from a plurality of financial institutions associated with a target user, a plurality of transactions associated with the target user. The computing system groups the plurality of transactions into a category or subcategory based on transaction data associated with each respective transaction. The computing system determines, based on the plurality of transactions, that the target user has a self-control problem. The computing system, responsive to determining that the target user has the self-control problem, generates an alert or recommendation to the target user. The alert or recommendation notifies the target user of the self-control problem.

Abstract: A processor may receive a call transcript including text and form a text string including at least a portion of the text. The processor may generate a situation description of the call transcript, which may comprise processing the text string using a transformer-based machine learning model. The processor may generate a trouble description of the call transcript, which may comprise creating a sentence embedding of the situation description, creating sentence embeddings for a plurality of utterances within the portion of the text, determining respective similarities between the sentence embedding of the situation description and each of the sentence embeddings for each respective one of the plurality of utterances, and selecting at least one of the plurality of utterances having at least one highest determined respective similarity as the trouble description. The processor may store a call summary comprising the situation description and the trouble description in a non-transitory memory.

Abstract: Systems and methods are used to generate contact type predictions that route user customer service requests within a support platform. The contact type predictions are generated using a hybrid model that includes a deep learning component and a business logic component. The deep learning component may generate a multi-channel output based on text features and context features. The multi-channel output is modified based on one or more business rules to generate the contact type predictions.