Abstract: A method for permitting a program to be analyzed and tuned includes the steps of receiving a program which itself includes a plurality of stages, where execution of the program follows a plurality of distinct paths, each of the distinct paths including one or more of the stages. The method further includes determining unique paths of the program and determining execution boundaries of the program. In addition, the method includes persisting the unique paths and the execution boundaries in a database, and changing an execution of the program based on the persisted unique paths and execution boundaries.

Abstract: An ad orchestration server selects and sequences the delivery of ads based on multiple offline and online conditions. The ad orchestration server takes into account messages a user might have seen from the advertiser online which the ad server knows about through its tracking cookie, as well as other data about the user that is obtained using offline and other online methods.

Abstract: Personalized messages are generated from a message template with high throughput by transforming data sources of the message template to key-value data structures that are stored in-memory and performing data look-ups using the in-memory key-value data structures.

Abstract: A method for permitting a program to be analyzed and tuned includes the steps of receiving a program which itself includes a plurality of stages, where execution of the program follows a plurality of distinct paths, each of the distinct paths including one or more of the stages. The method further includes determining unique paths of the program and determining execution boundaries of the program. In addition, the method includes persisting the unique paths and the execution boundaries in a database, and changing an execution of the program based on the persisted unique paths and execution boundaries.

Abstract: A method for updating records associated with work items within a business process workflow according to embodiments of the invention, includes the steps of processing the work items according to the business process workflow, and updating a record associated with a group of work items as the group of work items transitions from a first state to a second state within the business process workflow.

Abstract: A method for permitting a program to be analyzed and tuned includes the steps of receiving a program which itself includes a plurality of stages, where execution of the program follows a plurality of distinct paths, each of the distinct paths including one or more of the stages. The method further includes determining unique paths of the program and determining execution boundaries of the program. In addition, the method includes persisting the unique paths and the execution boundaries in a database, and changing an execution of the program based on the persisted unique paths and execution boundaries.

Abstract: A method for permitting a program to be analyzed and tuned includes the steps of receiving a program which itself includes a plurality of stages, where execution of the program follows a plurality of distinct paths, each of the distinct paths including one or more of the stages. The method further includes determining unique paths of the program and determining execution boundaries of the program. In addition, the method includes persisting the unique paths and the execution boundaries in a database, and changing an execution of the program based on the persisted unique paths and execution boundaries.

Abstract: A method for analyzing database records using sampling includes querying a recipient database for recipients with attributes that match one or more specified recipient attributes, querying a sampling database for transactions with attributes that match one or more specified transaction attributes, where the sampling database stores a subset of transactions that are also stored in a transaction database, intersecting the recipients resulting from the querying of the recipient database and the transactions resulting from the querying of the sampling database based on attributes thereof to create a set of joined sample results, and determining an approximate full count value by extrapolating a count value associated with the set of joined sample results based on both a total number of recipients in the recipient database and a total number of transactions in the transaction database.

Abstract: Personalized messages are generated from a message template with high throughput by transforming data sources of the message template to key-value data structures that are stored in-memory and performing data look-ups using the in-memory key-value data structures.