Abstract: Methods and systems for using machine learning to automatically determine a data loading configuration for a computer-based rule engine are presented. The computer-based rule engine is configured to use rules to evaluate incoming transaction requests. Data of various data types may be required by the rule engine when evaluating the incoming transaction requests. The data loading configuration specifies pre-loading data associated with at least a first data type and lazy-loading data associated with at least a second data type. Statistical data such as use rates and loading times associated with the various data types may be supplied to a machine learning module to determine a particular loading configuration for the various data types. The computer-based rule engine then loads data according to the data loading configuration when evaluating a subsequent transaction request.

Abstract: A system is disclosed for coordinating multiple software component deployments, upgrades, or migrations simultaneously or individually across a multiple-location/cloud platform. The system includes an application gateway router that routes incoming API requests based on an identifier in headers of the request that is associated with a swimlane construct, the swimlane containing multiple software components. By associating a swimlane's possible destinations with a timestamp or time range indicating when they are in effect, all of the components in the swimlane can be simultaneously upgraded or migrated by changing the API endpoint to which requests will be forwarded, with zero downtime. This solution is technology-, platform-, and cloud-agnostic and can be extended and applied to any organization using software.

Abstract: A device may receive assessment scores for a candidate associated with an entity and performance data identifying performance metrics and time periods associated with existing members of the entity. The device may process the assessment scores and the performance data, with an attrition model, to identify attrition scores for the candidate and confidences of the attrition scores, and may calculate a final attrition score based on the attrition scores. The device may process the assessment scores and the performance data, with a performance model, to identify performance scores for the candidate and confidences of the performance scores, and may calculate a final performance score based on the performance scores. The device may calculate an overall score based on the final attrition score and the final performance score, and may perform one or more actions based on the overall score.

Abstract: Methods and systems for using machine learning to automatically determine a data loading configuration for a computer-based rule engine are presented. The computer-based rule engine is configured to use rules to evaluate incoming transaction requests. Data of various data types may be required by the rule engine when evaluating the incoming transaction requests. The data loading configuration specifies pre-loading data associated with at least a first data type and lazy-loading data associated with at least a second data type. Statistical data such as use rates and loading times associated with the various data types may be supplied to a machine learning module to determine a particular loading configuration for the various data types. The computer-based rule engine then loads data according to the data loading configuration when evaluating a subsequent transaction request.

Abstract: A device may receive assessment scores for a candidate associated with an entity and performance data identifying performance metrics and time periods associated with existing members of the entity. The device may process the assessment scores and the performance data, with an attrition model, to identify attrition scores for the candidate and confidences of the attrition scores, and may calculate a final attrition score based on the attrition scores. The device may process the assessment scores and the performance data, with a performance model, to identify performance scores for the candidate and confidences of the performance scores, and may calculate a final performance score based on the performance scores. The device may calculate an overall score based on the final attrition score and the final performance score, and may perform one or more actions based on the overall score.

Abstract: Methods and systems for automatically discovering data types required by a computer-based rule engine for evaluating a transaction request are presented. Multiple potential paths for evaluating the transaction request according to the rule engine are determined. An abstract syntax tree may be generated based on the rule engine to determine the multiple potential paths. Based on an initial set of data extracted from the transaction request, one or more potential paths that are determined to be irrelevant to evaluating the transaction request are identified. Types of data required to evaluate the transaction request according to the remaining potential paths are determined. Only data that corresponds to the determined types of data is retrieved to evaluate the transaction request.

Abstract: A system for triage and response to a second system's malfunction is disclosed. The system determines a possible present or future disruption to a service provided by the second system; automatically searches historical records regarding similar disruptions; and provides an interactive, aggregated user interface comprising many tools to display results of the search, log actions already taken, and receive commands from human operators for seamless intervention in cloud-based or on-premises systems. The determination may be based on a statistical anomaly in the performance of the second system, or alternatively may be based on input from a human user, whose communication undergoes text analysis to determine possible matches.

Abstract: Methods and systems for using machine learning to automatically determine a data loading configuration for a computer-based rule engine are presented. The computer-based rule engine is configured to use rules to evaluate incoming transaction requests. Data of various data types may be required by the rule engine when evaluating the incoming transaction requests. The data loading configuration specifies pre-loading data associated with at least a first data type and lazy-loading data associated with at least a second data type. Statistical data such as use rates and loading times associated with the various data types may be supplied to a machine learning module to determine a particular loading configuration for the various data types. The computer-based rule engine then loads data according to the data loading configuration when evaluating a subsequent transaction request.

Abstract: A method of software version management for ensuring stability of published code. The method comprises receiving source code for an application; automatically inserting, into the source code or in a testing environment that will execute the source code, additional code to register an attempt by the source code to access an external resource; executing the source code in the testing environment; during execution of the source code, automatically tracking numbers of attempts made to access each external resource of a plurality of external resources; receiving configuration data associating each external resource with a fitness score; based at least in part on each number of attempts to access an external resource and on the fitness score associated with the external resource, determining a total application fitness score; and automatically publishing the source code to a production environment if and only if the total application fitness score exceeds a predetermined threshold.

Abstract: A method of software version management for ensuring stability of published code. The method comprises receiving source code for an application; automatically inserting, into the source code or in a testing environment that will execute the source code, additional code to register an attempt by the source code to access an external resource; executing the source code in the testing environment; during execution of the source code, automatically tracking numbers of attempts made to access each external resource of a plurality of external resources; receiving configuration data associating each external resource with a fitness score; based at least in part on each number of attempts to access an external resource and on the fitness score associated with the external resource, determining a total application fitness score; and automatically publishing the source code to a production environment if and only if the total application fitness score exceeds a predetermined threshold.

Abstract: Methods and systems for automatically discovering data types required by a computer-based rule engine for evaluating a transaction request are presented. Multiple potential paths for evaluating the transaction request according to the rule engine are determined. An abstract syntax tree may be generated based on the rule engine to determine the multiple potential paths. Based on an initial set of data extracted from the transaction request, one or more potential paths that are determined to be irrelevant to evaluating the transaction request are identified. Types of data required to evaluate the transaction request according to the remaining potential paths are determined. Only data that corresponds to the determined types of data is retrieved to evaluate the transaction request.

Abstract: Methods and systems for using machine learning to automatically determine a data loading configuration for a computer-based rule engine are presented. The computer-based rule engine is configured to use rules to evaluate incoming transaction requests. Data of various data types may be required by the rule engine when evaluating the incoming transaction requests. The data loading configuration specifies pre-loading data associated with at least a first data type and lazy-loading data associated with at least a second data type. Statistical data such as use rates and loading times associated with the various data types may be supplied to a machine learning module to determine a particular loading configuration for the various data types. The computer-based rule engine then loads data according to the data loading configuration when evaluating a subsequent transaction request.

Abstract: While website or other user content is often static, dynamic content personalization may allow for a better user experience. One way to personalize content, such as web content, is to gather information on a user, and then use that information to augment content. This approach is limited, however, in that a smaller website operator may have little or no data on its users, who may be infrequent. By making use of data collection modules, however, personalization data can be gathered from a number of different sources. Correlation of user identity using a correlation database may then be performed to determine that a user of one site is the same as the user of another site (even if the user presents different login credentials on those sites). This allows personalization data to be leveraged at scale and presentation of dynamic content opportunities, improving content and providing a more useful experience.

Abstract: While web site content is often static, dynamic web content personalization may allow for a more beneficial user experience. The present disclosure discusses gathering information on a user from a plurality of different websites, on which the user may be identified differently. Based on common linking information, a correlation can be made, however, between the different user identifiers on the different websites. With this correlation, potentially sparse information gathered from many different smaller websites can be aggregated and logically connected, enabling greater depth of knowledge on users' profiles. Profile data based on user actions on different websites (such as selecting an item for purchase) can be transmitted to other systems which can then use this data to present dynamic personalized content. In some embodiments, artificial intelligence techniques can further refine the usefulness of the profile data that is collected.