Abstract: Systems and methods that may implement an Oracle-aided protocol for producing and using FHE encrypted data. The systems and methods may initially encrypt and store input data in one encrypted form that is not performed using FHE, which does not substantially increase the size of the data and storage resources required to store the encrypted data. In accordance with the Oracle-aided protocol, the encrypted data is re-encrypted as FHE encrypted data when FHE encrypted data is required.

Abstract: A computing system receives, from a client device, a first request for access to a single page application associated with the computing system. A gateway of the computing system intercepts the first request before it reaches a shell service associated with the single page application. The gateway sends a second request for the single page application to the shell service. Based on the single page application request, the gateway determines that the shell service is experiencing a failure. Responsive to determining that the shell service is experiencing a failure, the gateway retrieves, from a content delivery network, a uniform resource locator (URL) associated with a time-lapse hypertext markup language (HTML) of the single page application. The gateway redirects the client device to the time-lapse HTML of the single page application using the URL.

Abstract: Aspects of the present disclosure provide techniques for integration scoring. Embodiments include importing a set of values from a data source to an electronic data system and assigning each value of the set of values to an import variable of a set of import variables in the electronic data system. Embodiments include displaying in a user interface a value for each import variable in the set of import variables in the electronic data system and receiving, via the user interface, one or more corrections to one or more import variables in the set of import variables. Embodiments include determining a weight corresponding to each import variable for which a correction was received. Embodiments include determining an integration score based on the one or more corrections and the weight corresponding to each import variable in the set of import variables for which a correction was received via the user interface.

Abstract: A computer-implemented method and system are provided to utilize machine learning technology to process user financial transaction data to predict a personalized payment screen architecture. A plurality of feature datasets associated with transaction data of a plurality of electronic invoices are obtained by a computing device. Each feature dataset comprises a plurality of features, a payment screen and a payment method configured to be presented on at least one payment screen. The computing device is configured to train a machine learning model with the feature datasets to produce a probability matrix with probabilities of each payment method used to pay the invoices through each payment screen. The computing device may weigh the probability matrix to generate a recommendation matrix and determine a prediction of a payment screen based on the recommendation matrix.

Abstract: A rule having text is pre-processed by replacing terms with dummy tokens. A first machine learning model (MLM) uses the dummy tokens to generate a dependency graph with nodes related by edges tagged with dependency tags. A second MLM uses the dependency graph to generate a canonical version with node labels. The node labels are sorted into a lexicographic order to form a document. A third MLM uses the document to generate a machine readable vector (MRV) that embeds the document as a sequence of numbers representative of a structure of the rule. The MRV is compared to additional MRVs corresponding to additional rules for which computer useable program code blocks have been generated. A set of MRVs is identified that match the MRV within a range. The set of MRVs correspond to a set of rules from the additional rules. The set of rules is displayed to a user.

Abstract: A method for computer estimations based on statistical tree structures involves obtaining a statistical tree structure for reference elements. The statistical tree structure includes leaf nodes segmenting a statistic for a data label according to data features in the reference elements, and intermediate nodes connecting a first node to the leaf nodes. Each of the first node and the intermediate nodes provide a branching based on one of the data features. The method further includes obtaining target data, including values for the data features, and a value for the data label. The method also includes selecting the first node, associated with a first data feature, traversing the statistical tree structure to a leaf node by matching the values of the data features to the branching of the intermediate nodes, and assessing the value for the data label in the target data based on the statistic associated with the leaf node.

Abstract: Big data analysis methods and machine learning based models are used to provide offer recommendations to consumers that are probabilistically determined to be relevant to a given consumer. Machine learning based matching of user attributes and offer attributes is first performed to identify potentially relevant offers for a given consumer. A de-duplication process is then used to identify and eliminate any offers represented in the offer data that the consumer has already seen, has historically shown no interest in, has already accepted, that are directed to product or service types the user/consumer already owns, for which the user does not qualify, or that are otherwise deemed to be irrelevant to the consumer.

Abstract: Aspects of the present disclosure provide techniques for document classification through signal processing. Embodiments include receiving a document for classification. Embodiments include generating an image of the document. Embodiments include producing a signal representation of the document based on numbers of non-white pixels in each horizontal scan line or vertical scan line of the image of the document. Embodiments include comparing the signal representation of the document to signal representations of previously-classified documents. Embodiments include determining, based on the comparing, a classification for the document. Embodiments include performing additional processing with respect to the document based on the classification for the document.

Abstract: A method may include executing a baseline classifier on unreviewed transaction features of an unreviewed transaction record to obtain a baseline account identifier, and executing a comparison model on (i) an unreviewed transaction vector of the unreviewed transaction record and (ii) reviewed transaction vectors to obtain comparison scores. The reviewed transaction vectors may correspond to reviewed transaction records each having a user-approved account identifier. The method may further include selecting, using the comparison scores, a reviewed transaction record. The reviewed transaction record may correspond to a comparison score. The comparison score may correspond to a user-approved account identifier of the reviewed transaction record.

Abstract: Systems and methods are provided for categorizing a transaction description using natural language processing.

Abstract: A server computer may receive and process a plurality of time series data to generate sparse datasets based on sparsity levels. The server computer applies a time series forecasting model to each respective subset of previous data points of the sparse datasets increasingly at the first time granularity to generate a set of prediction values and a set of residuals; applies a regression model to the set of the prediction residuals to generate a set of adjusted residuals for the sparse datasets; and generates a visualized explanation based on the set of the prediction values and the set of adjusted residuals for one or more of the sparse datasets.

Abstract: Systems and methods for recommending and automating user savings goals within a single banking account are disclosed. An example method is performed by an electronic device coupled to a machine learning model and includes training the machine learning model based at least in part on historical user data, receiving attributes of a first user, generating, using the trained machine learning model, one or more recommended savings goals for the first user based at least in part on the attributes of the first user, receiving selection of one or more of the recommended savings goals, determining, using a trained classification model, a recommended amount for the first user to periodically save, the recommended amount determined based at least in part on the attributes of the first user, and periodically allocating savings to each of the one or more selected savings goals based at least in part on the recommended amount.

Abstract: A method for unbiased interviewing involves, during an interview of an interviewee: capturing an interviewee utterance, generating a voice-modulated interviewee utterance from the interviewee utterance, transmitting the voice-modulated interviewee utterance to an interviewer processing system, capturing an interviewee image, and transmitting the interviewee image to a backend processing system.

Abstract: This disclosure relates to automatic text information extraction from electronic documents. An example system is configured to perform operations including obtaining text and one or more document features of an electronic document, clustering the text into one or more groups based on the one or more document features, and identifying one or more text strings from the text in one or more groups as one or more keys. Identifying the one or more text strings is based on the clustering. The system is also configured to perform operations including generating one or more key/value pairs. Generating one or more key/value pairs includes associating one or more values to the one or more keys (with a value including text outside of the one or more identified text strings). The system is further configured to output the one or more key/value pairs.

Abstract: Systems and methods for controlling traffic flow for software experimentation are disclosed. An example method includes receiving a first traffic from a first plurality of users to receive an existing feature or a new feature, determining, from the first traffic, a first portion of the first plurality of users to receive the existing feature (with the first portion bounded by a first lower threshold and/or a first upper threshold) and a second portion of the first plurality of users to receive a new feature (with the second portion bounded by a second lower threshold and/or a second upper threshold). The method also includes providing the existing feature to the first portion, providing the new feature to the second portion, receiving feedback from the first plurality of users, and adjusting the first portion and/or the second portion based on the feedback and based on the bounds.

Abstract: A method may include receiving, from a browser, a request for a tracking pixel including a uniform resource locator (URL) of the tracking pixel and a session identifier (ID) of the browser. The URL of the tracking pixel may be included in a CSS selector. The method may further include sending, to the browser and as triggered by the request for the tracking pixel, a request to scan a document object model (DOM) of the browser. The request to scan the DOM may include the session ID. The method may further include receiving, from the browser, scan results including a set of dependencies on the CSS selector.

Abstract: Certain aspects of the present disclosure provide techniques for providing detailed customer-level lifetime value data via a user interface. Embodiments include receiving customer data related to a customer. Embodiments include using a plurality of predictive models to predict, based on the customer data, values for a plurality of sub-components of a lifetime value of the customer, wherein each predictive model of the plurality of predictive models corresponds to a sub-component of the sub-components. Embodiments include determining customer-level lifetime value data for the customer by aggregating the values for the plurality of sub-components. Embodiments include providing the customer-level lifetime value data for the customer, including a subset of the values for the sub-components, to an application for display to a user via the user interface.

Abstract: Aspects of the present disclosure provide techniques for explainable payroll calculations. Embodiments include receiving a request from a client for a rule-based calculation. Embodiments include receiving one or more user values related to the request. Embodiments include using a calculation graph to determine the rule-based calculation based on the one or more user values. The calculation graph may comprise at least one node that performs an operation using the one or more user values. Embodiments include identifying an explanation template associated with the at least one node. Embodiments include generating an explanation of the rule-based calculation based on the explanation template and the one or more user values. Embodiments include providing the rule-based calculation and the explanation of the rule-based calculation to the client in response to the request.

Abstract: A task execution application programming interface may include a pipeline execution service interface configured to communicate between a task image and a pipeline execution service adapter. The pipeline execution service adapter may be configured to receive, from a pipeline execution service, a request to execute the task image in a pipeline. The request may include arguments. The task image may include executable code and an execution environment. The pipeline execution service interface may be further configured to obtain results by executing the executable code using the arguments in the execution environment. The pipeline execution service adapter may be further configured to provide, to the pipeline execution service, access to the results. The pipeline execution service may control execution of the pipeline using the results.

Abstract: Disclosed methods and systems provide domain-specific type ahead suggestions for search query terms with a customer self-service system for a tax return preparation system, according to one embodiment. Type ahead suggestions include suggestions for completing a single search query term and/or suggestions for completing a search query having multiple search query terms. The customer self-service system receives one or more search query terms in a search text box, searches for potential search queries that are likely to complete the one or more search query terms, and provides a selected one of the potential search queries in the search text box as a type ahead suggestion, adjacent to the received one or more search query terms, according to one embodiment. The customer self-service system selects one of the potential search queries based on the received search query terms, the user's profile, and/or the user's tax data, according to one embodiment.