Abstract: Reduction in latency between question submissions and response submissions in a question and answer based customer support system is reduced by facilitating the use of mobile devices by customer support personnel to submit question responses. The answerability of a question from a mobile device is predicted by analyzing the attributes of the question itself before the answer is generated. Questions being entered into the question and answer based customer support system that are conveniently answerable from a mobile device are routed to a mobile question and answer queue that enables mobile device users to review and respond to the mobile device answerable questions. A question is determined to be answerable from a mobile device if a predicted answer to the question has a length that is within a predetermined threshold, so that customer support personnel are not burdened with generating long responses from mobile devices.

Abstract: Certain aspects of the present disclosure provide techniques for providing personalized groups of travel data for review through a user interface. Embodiments include receiving trip records associated with a user from an application running on a remote device, providing the trip records to a prediction model, and receiving a plurality of groups from the prediction model, each group of the plurality of groups comprising a subset of the trip records. Embodiments include providing each group of the plurality of groups to a personalization model, the personalization model having been trained based on user feedback to determine personalization scores for each group of the plurality of groups. Embodiments include receiving a personalization score for each group of the plurality of groups from the personalization model and transmitting one or more groups selected based on the personalization scores to the application to be displayed via the user interface.

Abstract: Method and apparatus for providing personalized self-help experience in online application. A predictive model is trained to learn a relationship between one or more user features and one or more tags using historical user feature data. High-dimensional vectors representing each of a plurality of questions are generated and stored in the lookup table. The trained predictive model outputs tags probabilities from the incoming user data, using the learned relationship. A user high-dimensional vector is formed based on the tags probabilities. Similarity metrics are calculated between the high-dimensional vector for the respective question and the user high dimensional vector. One or more of the most relevant question titles are returned to a client device for presentation to a user.

Abstract: An encrypted database system includes a memory storing a database comprising a plurality of logical structural elements each respectively including an unencrypted fuzzed value and encrypted sensitive data formed by encrypting a sensitive data value. The system also includes a processor in communication with the memory and configured to form the plurality of logical structural elements and store the plurality of logical structural elements in the memory. Forming a logical structural element comprises generating the unencrypted fuzzed value for the sensitive data value, encrypting the sensitive data value, and storing the encrypted sensitive data value and the unencrypted fuzzed value in the same logical structural element in the database. The unencrypted fuzzed value is within a predetermined value range and is different from the sensitive data value.

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: Embodiments presented herein provide a partner authentication (PA) system that coordinates a network-based authorization process for an application. The PA system exchanges a series of messages with the application seeking an access token for a protected resource, an authorization server associated with the resource, and an agent executing on a device accessed by a user who wants the application to access the resource. The PA system and the agent communicate with the authorization server on behalf of the application throughout the authorization process. At the completion of the authorization process, the PA system receives an access token and a refresh token from the server on behalf of the application and sends a partner authorization (PA) token to the application. When the application seeks access to the resource that is available to authorized parties via the resource server, the application sends the PA token to the PA system and receives the access token in return.

Abstract: Systems and methods for identifying solicitations of personally identifiable information (PII) are disclosed. An example method includes generating training data based on historical transcript data corresponding to solicitations of PII in historical support call transcripts, training a neural network, using the training data, to identify solicitations of PII in support call transcripts, and processing the trained neural network for deployment.

Abstract: A method and system provides a data management system that provides data management services and products to users. The method and system provides a predictive model that generates probability scores indicating the likelihood that current users of the data management system would select promotional messages if the promotional messages are presented to the current users.

Abstract: Certain aspects of the present disclosure provide techniques for determining crash similarity based on stack traces and user actions associated with a crash event and handling crash events based on the determined crash similarity. An example method generally includes receiving a crash report for a crash event recorded during execution of a software application. A similarity score is generated between the received crash report and one or more historical crash reports associated with historical crash events recorded during execution of the software application. A target set of historical crash reports is selected based on the generated similarity scores for each respective historical crash report. The target set of historical crash reports generally includes crash reports having generated similarity scores exceeding a threshold similarity score. One or more actions are taken to process the received crash report based on the target set of historical crash reports.

Abstract: Systems and methods for automatically identifying problem-relevant sentences in a transcript are disclosed. In an example method, a transcript may be received of a first support call. A region of the first support call transcript may be identified, and first customer utterances may be detected in the region using a trained classification model. A trained regression model may estimate a relevancy to the problem statement of each of the first customer utterances, and one or more most problem-relevant statements may be selected from the first customer utterances, based on the estimated relevancies.

Abstract: Systems and methods for forecasting cashflows across one or more accounts of a user disclosed. One example method may include retrieving a data set for each of a plurality of accounts from a database, constructing a graph including a plurality of nodes linked together by a multitude of edges, wherein each node identifies a time series value corresponding to one of the accounts, and each edge indicates a time series value of a corresponding set of transactions occurring between a corresponding pair of accounts, determining a plurality of constraints, determining a specified loss function based on the plurality of constraints, back-propagating a derivative of the specified loss function into a deep neural network (DNN) to determine a set of neural network parameters, forecasting, using the DNN, a time sequence for one or more of the nodes and one or more of the edges, and providing the forecasted time sequences to the user.

Abstract: A method may include registering, with an offline job to be executed by a computer processor, an application programming interface (API) and an operation, obtaining, from a repository, a user consent of a user for the operation, and in response to obtaining the user consent, creating, for the user, an access token including the operation and the API. The user consent may be stored external to the access token. The method may further include transmitting the access token to the offline job, and calling, by the offline job, the API using the access token.

Abstract: Dialog systems, such as chatbots, enable a user to query for information of a data management system. However, a user's query may be ambiguous resulting in a reply of too much information that is incomprehensible to a user. Clarifying questions are posed to a user to remove ambiguity from a user's query, and based on a user's answers, the volume of replying information is reduced to an amount that a user can comprehend. Clarifying questions are efficiently sequenced to reduce the ambiguity of a user's query, such that the number of clarifying questions does not result in a user abandoning the user's query.

Abstract: Systems and methods for call handling optimization of a call center are disclosed. A system is configured to: obtain a plurality of agent demand variables; obtain an indication of available agents that are associated with a plurality of skills; generate a multi-skill routing matrix depicting a routing problem from the agent demand variables and the available agents; deconstruct the multi-skill matrix into one or more band routing matrices; and identify, for each band routing matrix, a number of agents from the available agents to be used in solving the band routing matrix. Identifying the number of agents includes recursively simulating a plurality of possible routing solutions, measuring a quality metric for each simulation, and reducing the range of available agents based on the quality metric. The system may indicate the number of agents to be used for each routing sub-problem (which may be at the staff group level).

Abstract: A method for facilitating user support using multimodal information involves obtaining an interaction between a user and a support agent, generating a question embedding from the interaction, obtaining a clickstream associated with the interaction, and generating a clickstream embedding from the clickstream. The question embedding and the clickstream embedding form a shared latent space representation. The method further involves decoding a problem summary from the shared latent space representation and providing the problem summary to the support agent.

Abstract: A processor may filter data to generate a subset of the data less than an entire set of the data. The subset may include at least one string and at least one numeric value. The processor may match the at least one string and the at least one numeric value to one of a plurality of archetypes by applying a clustering algorithm. Each archetype may include a subset of archetype data less than an entire set of archetype data. The processor may compare the entire set of data to the entire set of archetype data to identify at least one difference between the entire set of data and the entire set of archetype data. The processor may apply at least one optimization to address the at least one difference.

Abstract: Disclosed methods and systems improve search results by recency boosting customer support content for a customer self-help system associated with one or more financial management systems. The customer self-help system retrieves content relevance from a variety of sources, such as media outlets, taxation agencies and news feeds for the financial management system. The customer self-help system generates content relevance weights from the content relevance data, and applies the content relevance weights to customer support content maintained by the customer self-help system. In response to receiving a search query from a user, the customer self-help system provides relevant portions of customer support content that has been recency boosted (e.g., adjusted by the content relevance weights), to increase the likelihood that the customer support content provided to the user is relevant to the user's search query.

Abstract: Systems and methods for automatically assessing a quality of service for agents of a customer support system are disclosed. An example method may include retrieving historical conversations between the agents and users of the customer support system, receiving user comments for one or more of the historical conversations, identifying which of the received user comments includes keywords indicative of one or more quality of service attributes, generating transcripts of historical conversations associated with the identified user comments, training a machine learning model based at least in part on the generated transcripts and the user comments of the historical conversations associated with the identified user comments, providing a plurality of current conversations between agents and users of the customer support system to the trained machine learning model, and generating a behavioral score for each of the agents using the trained machine learning model.

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.