Abstract: Described herein are computerized methods and systems for user authentication using an imaged machine-readable identity document. A server receives an authentication request from a first client device, including image files corresponding to a user's machine-readable identity document. The server displays on the first client device user-identifying data elements extracted from the image files. The server captures additional user-identifying data elements from the first client device, and verifies the user's identity based upon the user-identifying data elements. The server determines user contact channel data based upon the verified identity. The server displays the contact channel data on the first client device, and generates a transient access code upon receiving a contact channel selection.

Abstract: VoC data, e.g., plurality of VoC communications, for a plurality of customers can be collected. A clustering algorithm may be utilized to generate a plurality of cluster each of which is assigned a group of customers that share similar metadata. An emotion profile may be generated for each cluster based on segment detected emotions for the VoC data corresponding to the customers that are assigned to the cluster. The emotion profile for a cluster may represent the normal or expected emotions experienced during VoC communications by customers assigned to the cluster. New VoC data corresponding to a new customer may be received. A cluster may be identified based on the metadata of the new customer, and the emotion attributes of the new VoC data may be compared to the emotion profile of the identified cluster. Based on the comparison, one or more deviation can be identified to trigger an action.

Abstract: Methods and apparatuses are described for enhancing chatbot recognition of user intent through graph analysis. A computing device classifies user messages associated with a chat session as either comprehensible user messages that are mapped to a user intent or incomprehensible utterances that are not mapped to a user intent. The computing device arranges the comprehensible messages into first clusters and arranges the incomprehensible messages into second clusters. The computing device generates a first sub-graph for each first cluster and generates a second sub-graph for each second cluster. The computing device calculates a graph kernel matrix using the sub-graphs and trains a classification model using the graph kernel matrix. The computing device executes the trained classification model using the graph kernel matrix as input to generate a predicted intent identifier for the incomprehensible messages.

Abstract: Methods and apparatuses are described for capturing and maintaining a timeline of data changes in a relational database system. A server identifies changed records from relational database tables. The server analyzes the changed records to determine a maximum timestamp for each primary key and extracts the changed records associated with each primary key where a timestamp is equal to or greater than the maximum timestamp for the primary key. The server generates timestamp ranges for each primary key, each comprising an effective date and an expiration date. The server determines whether each key-date combination already exists in a historical record table. The server updates an expiration date of an existing record in the historical record table using the effective date and inserts a new record for the timestamp range using the captured records.

Abstract: Computerized systems and methods are provided for visualizing data in a virtual environment. A three-dimensional virtual display framework is provided within a three-dimensional virtual space, the virtual display framework defined by a three-dimensional volumetric display zone bound by a plurality of surfaces. Data is visualized in 3D within the volumetric display zone of the virtual display framework. Select information is also displayed on at least one of the surfaces of the virtual display framework, where the select information is relevant to the data visualized. In addition, one or more display parameters can be altered in response to hand motions of the user captured to customize display of the virtual display framework. Altering the display parameters is adapted to adjust at least one of the three-dimensional visualization of the data within the volumetric display zone or the display of the select information on the at least one surface.

Abstract: Methods and apparatuses are described in which unstructured computer text is analyzed for generation of customized digital documents. A server tokenizes and encodes historical user interactions and historical digital documents into multidimensional vectors. The server trains an interaction classification model using the multidimensional vectors as input to generate a classification for an input user interaction, and trains a language generation model using the multidimensional vectors as input to generate a customized digital document based upon an input user interaction. The server receives a new user interaction and encodes the new user interaction into a new multidimensional vector. The server executes the trained interaction classification model using the new vector as input to generate a digital document classification. The server executes the trained language generation model using the new vector and the classification as input to generate a customized digital document.

Abstract: The Multichannel Master Feeder Exchange Mechanism Apparatuses, Methods and Systems (“MMFEM”) transforms user share purchase, redemption request inputs via MMFEM components into fund share transaction records outputs. A processor-implemented non-transparent, master-feeder, master-feeder investment vehicle management, multichannel datastructure apparatus, comprising a processor and memory. The MMFEM memory having instructions to instantiate a multichannel asset vehicle configuration datapath and create a the multichannel vehicle data structure. The MMFEM may then determine a plurality of investment share data structures for the multichannel asset vehicle data structure and facilitate a transaction of investment shares data structures of the multichannel investment vehicle data structure based on the determined share price, wherein a master fund establishes a data feed for feeder funds to track.

Abstract: Methods and apparatuses are described in which unstructured computer text is analyzed for generation of an enforcement action database. A computing device receives a digital document comprising a corpus of unstructured text and converts each sentence into tokens. The computing device identifies entities in the tokens and classifies each sentence as relevant or not relevant to an enforcement action. For each relevant sentence, the computing device generates a record in a first data structure and generates a record in a second data structure. The computing device maps the records in the first data structure to the records in the second data structure based upon an enforcement attribute to generate a third data structure. The computing device aggregates the records in the third data structure based upon a name of the entity and a type of the entity in the sentence to determine an aggregated first enforcement attribute for each entity.

Abstract: The Multiple Modular Asset Class Constructor Apparatuses, Methods and Systems (“MMACC”) transforms collateralized equity obligation structure parameters, asset search, tranche selections inputs via MMACC components into asset income distribution message, principal distribution message outputs. A capital structure input may be obtained from a system user via a first data structure that specifies dividend allocation and capital allocation for a first equity tranche, and dividend allocation and capital allocation for a second equity tranche. Dividend allocation and capital allocation for the first equity tranche, and dividend allocation and capital allocation for the second equity tranche may be determined. A second data structure may be generated that maps dividend allocation and capital allocation of an index to the first equity tranche and to the second equity tranche. Shares of the first equity tranche and shares of the second equity tranche may be generated using the second data structure.

Abstract: Methods and apparatuses are described for automatic refinement of intent classification for virtual assistant applications. A computing device creates utterance groups comprising messages. The device determines, using a first machine learning (ML) classifier, a first predicted intent associated with each utterance group and determines, using a second ML classifier, a second predicted intent associated with each utterance group. The device combines the first and second predicted intents to generate a final predicted intent, determines, for each incomprehensible utterance group, whether the final predicted intent overlaps with the final predicted intent for other utterance groups, and selects messages having no overlapping intent to create new intents.

Abstract: A computer-implemented method for controlling access to a source code repository residing on a server system coupled to a communications network includes: receiving, by a computing device coupled to the network, an instruction to execute a client executable for accessing the source code repository, wherein the client executable includes a first hash based on the client executable; determining a validity of the first hash; computing a second hash based on the client executable; providing the first hash and the second hash to the server system; determining a validation result for the client executable, wherein the validation result is based on the first hash and the second hash matching a corresponding pair of hashes of a plurality of valid hashes stored in a database; in response to the validation result indicating that the client executable is valid, accessing the source code repository.

Abstract: Methods and apparatuses are described for automated optimization and personalization of customer-specific communication channels using feature classification. A server captures historical interaction data comprising a channel type, a user identifier, an interaction date, and a user response value. The server generates a channel feature vector for each combination of channel type, user identifier, and interaction date. The server identifies features from the channel feature vectors for each different channel type and aggregates the features into a common feature vector. The server executes a trained classification model on the common feature vectors to select user identifiers for each different channel type that have an engagement probability value at or above a corresponding threshold.

Abstract: Techniques are provided for performing information technology service management (ITSM) correlation for infrastructure environment functions such as management, operation, and remediation. Specifically, a processor may receive an incident identifier for an incident record corresponding to an incident encountered by the infrastructure environment. The processor may use the incident identifier to correlate incident data to data of different management services, e.g., change management and problem management. Specifically, the processor may correlate the incident data to change data, and the change data may be correlated to problem data. The processor may also correlate the incident data to other incident data, and the other incident data may be correlated to problem data. The problem data can be utilized to improve or remediate the incident of interest.

Abstract: Methods and apparatuses are described for integrating common and context-specific natural language understanding (NLU) processing in a virtual assistant application. A server identifies a common NLU processing unit comprising (i) a common NLU classification model and (ii) a common response library comprising common responses. The server builds a combined NLU processing unit based upon the common NLU classification model and a context-specific NLU classification model, the combined NLU processing unit comprising (i) a combined NLU classification model trained using the common NLU training data and context-specific NLU training data, (ii) the common response library and (iii) a context-specific response library comprising context-specific responses. The server generates a response to an input utterance by determining a user intent associated with the input utterance using the combined NLU classification model, locating a response associated with the user intent in one or more of the libraries.

Abstract: In one embodiment, a method for accounting for equity transactions for an incorporated legal entity in an interactive computer system is disclosed, comprising: identifying a stock option grant document stored as an electronic document on an interactive computer system, the electronic document stored in association with an electronic record of the incorporated legal entity; identifying a stock option grant in the stock option grant document, stored as core record metadata within the stock option grant document; determining a number of granted stock options using the stock option grant stored as core record metadata; subtracting the number of granted options from a stock option grant user account associated with both a user and the incorporated legal entity; and subtracting the number of granted options from an outstanding shares account associated with the incorporated legal entity, in conjunction with a stock option exercise document and as a number of granted shares.

Abstract: Methods and apparatuses are described for determining future user actions using time-based featurization of clickstream data. A server captures clickstream data corresponding to web browsing sessions and converts the clickstream data into tokens by identifying each unique URL and parsing each unique URL into tokens. The server generates a frequency matrix based upon the tokens, and generates a latent feature vector for each URL in the session based upon the frequency matrix. The server merges the latent feature vectors and the clickstream data into an aggregate clickstream vector set for a user. The server assigns time-decayed weight values to each latent feature vector in the aggregate clickstream vector set. The server combines the time-decayed latent feature vectors to generate a clickstream embedding for the user, and executes a machine learning model using the clickstream embedding to generate one or more predicted actions of the user.

Abstract: Systems and methods for extracting data values from electronic documents using neural networks. The method includes receiving an electronic document having data values and associated field identifiers, determining pixel coordinates corresponding to the field identifiers using a first neural network, and extracting the field identifiers located at the pixel coordinates using a second neural network. The method also includes, for each of the field identifiers, calculating pixel coordinates on the electronic document corresponding to a data value associated with the field identifier using a third neural network and extracting the data value located at the calculated pixel coordinates using the second neural network. The method further includes, for each of the data values, generating a record in a data structure, the record including the extracted value and the extracted field identifier. The method also includes storing the data structure including the records in a database.

Abstract: Methods and apparatuses are described for automated prediction of computing resource performance scaling using reinforcement learning. A server executes performance tests against a production computing environment comprising a plurality of computing layers to capture performance data for computing resources in the production environment, where the performance tests are configured according to transactions-per-second (TPS) values. The server trains a classification model using the performance data, the trained model configured to predict computing power required by the plurality of computing layers. The server identifies a target TPS value and a target cost tolerance for the production environment and executes the trained classification model using the target TPS value and the target cost tolerance as input to generate a prediction of computing power required by the plurality of computing layers.

Abstract: Systems and methods are described herein for stopping a data replication process on database systems. The method includes identifying a data replication process corresponding to a downstream database system and an upstream database system. The method also includes determining a first status of the data replication process at the downstream database system and upstream data corresponding to the upstream database system. The method also includes determining a second status of the data replication process at the upstream database system based on the upstream data. The method also includes stopping the data replication process at the upstream database system based on the second status and stopping the data replication process at the downstream database system based on the first status. The method also includes generating a data log notification including data corresponding to the stopped data replication process.

Abstract: Methods and apparatuses are described for provisioning and managing data orchestration platforms in a cloud computing environment. A server provisions in a first region a first data orchestration platform comprising (i) a first data transformation instance, (ii) first endpoints, and (iii) a first data integration instance. The server provisions in a second region a second data orchestration platform comprising (i) a second data transformation instance, (ii) second endpoints, and (iii) a second data integration instance. The server integrates the first data integration instance and the second data integration instance with an identity authentication service. The server monitors operational status of the first orchestration platform and the second orchestration platform using a monitoring service. The server refreshes virtual computing resources in each of the first orchestration platform and the second orchestration platform using a rehydration service.