A development engine may enable a developer to customize a user experience using an intuitive developer interface. A rules engine may provide constructs to a card engine in the form of card definitions, which the card engine may evaluate using facts obtained from a facts controller. The evaluated card definitions are cards that may be output for presentation via user equipment. Variants may be assigned weights which can be set or changed dynamically by the card engine substantively and in real time based on factors such as user behavior, account condition, promotions or offerings. The card engine may make content decisions proximate to events occurring to the user. The presentation of the cards may be changed substantively and in real time in accordance with the setting or changes in variants.

Runtime errors in a source code program are detected in advance of execution by machine learning models. Features representing a context of a runtime error are extracted from source code programs to train a machine learning model, such as a random forest classifier, to predict the likelihood that a code snippet has a particular type of runtime error. The features are extracted from a syntax-type tree representation of each method in a program. A model is generated for distinct runtime errors, such as arithmetic overflow, and conditionally uninitialized variables.

A method for instrumenting an update to a software application may include determining, based on a source code file affected by the update to the software application, a first method affected by the update to the software application. A second method called by the first method and a third method called by the second method may also be identified as being affected by the update to the software application. A user interface file that includes a call to the first method, the second method, and/or the third method may be identified. The functional impact of the update may be determined by identifying one or more functional flows that match the user interface file. A recommendation identifying the one or more matching functional flows as candidates for testing may be generated. Related systems and computer program products are also provided.

Methods, apparatuses, and computer program products are described. A system may run a set of Automated Structure processing (ASP) scripts on a program binary executable written in C-language without making changes to source-code. The system may generate a set of dictionary files indicating data structural information specifically for the C-language program based on running the set of ASP scripts on the program binary executable. The dictionary files may include a list of C-structures referenced and type definitions and structure layouts associated with the C-structures. The system may combine the data structural information with an ASP library to generate an extended ASP library, and may augment the program binary executable with the extended ASP library to generate an augmented program binary executable. The system may programmatically perform an introspection and data discovery procedure on a C-language program at run-time using the augmented program binary executable.

Disclosed herein are systems and method for detecting coroutines. A method may include: identifying an application running on a computing device, wherein the application includes a plurality of coroutines; determining an address of a common entry point for coroutines, wherein the common entry point is found in a memory of the application; identifying, using an injected code, at least one stack trace entry for the common entry point; detecting coroutine context data based on the at least one stack trace entry; adding an identifier of a coroutine associated with the coroutine context data to a list of detected coroutines; and storing the list of detected coroutines in target process memory associated with the application.

Techniques for evaluating software systems are provided, including measurement of performance ratings of features of an app-under-test are described. The evaluation may include analysis of a video recording of the user interface of the app-under-test, where the video analysis may include comparing the changes over time of co-located pixel in the video to produce a performance rating, for example a speed index. The results of multiple such tests can be compared by comparing the performance rating produced during execution of different tests.

Methods, systems, and computer-readable media for GPU code injection to summarize machine learning training data are disclosed. Training of a machine learning model is initiated using a graphics processing unit (GPU) associated with a machine learning training cluster. The training of the machine learning model generates tensor data in a memory of the GPU. The GPU determines a summary of the tensor data according to a reduction operator. The summary is smaller in size than the tensor data and is output by the GPU. A machine learning analysis system performs an analysis of the training of the machine learning model based at least in part on the summary of the tensor data. The machine learning analysis system detects one or more conditions associated with the training of the machine learning model based at least in part on the analysis.

Techniques are disclosed for setting a breakpoint for debugging a neural network. User input is received by a debugger program executable by a host processor indicating a target layer of a neural network at which to halt execution of the neural network. The neural network includes a first set of instructions to be executed by a first execution engine and a second set of instructions to be executed by a second execution engine. A first halt point is set within the first set of instructions and a second halt point is set within the second set of instructions. It is then determined that operation of the first execution engine and the second execution engine has halted. It is then determined that the first execution engine has reached the first halt point. The second execution engine is then caused to move through instructions until reaching the second halt point.

A computer implemented method includes accessing performance trace data for executed code of multiple services. Symbols corresponding to functions of the executed code are identified. First sequences of functions from the identified symbols are identified and a first performance threshold for each identified first sequence of functions is computed. The method includes receiving an incoming performance trace, detecting second sequences of functions from the incoming performance trace, identifying second sequences equivalent to the first sequences, and comparing performance of the identified second sequences to the first performance threshold for each of the equivalent first sequences to identify second sequences as comprising a performance bottleneck.

Embodiments disclose a method and a device for processing frequency converter monitoring data and a storage medium. An embodiment of the method includes: acquiring a script file containing a monitoring parameter specifying field and a storage location; parsing the script file to acquire the storage location and the monitoring parameter specifying field and determining the monitoring parameter specified by the monitoring parameter specifying field; collecting the monitoring data corresponding to the monitoring parameter; and storing the monitoring data in the storage location. The embodiments provide a script file based processing solution for frequency converter monitoring data, without no special tracking or debugging software tool required. Thus, the implementation complexity is reduced and the service cost and time are saved.