Systems and methods decouple model components from a model execution style for which the model components are created, and the model components may be utilized in parent models having different execution styles. A model component may be partitioned into executable entities, and the entry points of the executable entities and their call styles may be identified. An adaptation layer that includes access points for the entry points may be constructed. The model component, including the adaptation layer, may be included in the model, and connection elements of the parent model may be connected to the access points of the adaptation layer. The execution call styles associated with the connection elements of the parent model may be bound to the execution call styles of the entry points as originally designed. The adaptation layer may manage translation of call styles and may coordinate scheduling of data communication with the model component.

A computer-implemented method comprises obtaining a first build task for building first source code in a first programming language of a plurality of programming languages; retrieving, by the processor, the first source code based on the first build task; building the first source code into one or more artifacts and one or more job specifications; storing the one or more artifacts in a cache shared across a cluster; and initializing an application module on the cluster based on the first programming language, the application module configured to receive a job specification of the one or more job specifications and execute a data transformation job using a reference to a location in the cache.

A cross domain generalization system for industrial artificial intelligence (AI) applications is disclosed. A target encoder subsystem obtains target data from a target machine product and generates lower dimensional data for obtained target data using a target artificial intelligence (AI) model. The generated lower dimensional data are corresponding to a plurality of target embeddings data. The target encoder subsystem further applies the plurality of target embeddings data into a source classifier AI model. A source classifier subsystem predicts a quality of the target machine product by generating class labels for each of the plurality of target embeddings data based on a result of the classifier AI model. The goal of the present invention is to learn features or representations such that the correlation with a label space is similar both in source and target domains while being invariant of data distributions.

In one embodiment, a computing system identifies one or more terminal nodes in a digraph, and then back-walks primitives up the digraph from each terminal node to a corresponding parent terminal node or a root of the digraph, whichever is first. The system then identifies chains of back-walked primitives for each of the one or more terminal nodes (e.g., where each chain consists of a respective terminal node and any primitives either a) up to but not including a corresponding parent terminal node or else b) up to and including the root of the digraph when the back-walking reaches the root). Based on this, the system can then merge each set of any two or more chains of the identified chains that intersect on a decision operation into a corresponding new single chain, and maps each of the chains to a respective rule.

Functionality is provided for the automated creation, testing, training, adaptation and deployment of AI models and changes thereto. Base classes are provided that enable practicable creation of new models from existing one. New models are tested on live data sets offline from user sites. New training methods are provided for the production of particular outcomes. Efficient adaptation of new AI models is facilitated, encompassing data scientist and development team control over how fast to train and deploy new models.

An apparatus duplicates a process code of a variadic template function that has a variable number of parameters in a source code, in association with each of actual arguments in an actual-argument list corresponding to a variadic parameter defined by a variadic operator that packs the variable number of parameters of the variadic template function. The apparatus substitutes another parameter in each duplicated process code with a prepared parameter that accepts the actual argument associated with the each duplicated process code. The apparatus firstly inserts, into a recursive call part in a process code of the variadic template function, a first duplicated process code that is associated with an actual argument at a head of the actual-argument list, and repeats inserting, into a recursive call part in the previously inserted duplicated process code, a next duplicated process code associated with a subsequent actual argument.

Certain aspects of the disclosure are directed to programming of a data-communications system. According to a specific example, a data-communications (e.g., VoIP) server is configured to identify, in response to received calls from endpoint devices, a set of scripts written in a programming language that includes call flow commands. The server is further configured to execute the set of scripts to retrieve data from the data sources and control, in response to the data, call flow for the calls.

Source code of a graph analysis program expressed in a platform-independent language which supports linear algebra primitives is obtained. An executable version of the program is generated, which includes an invocation of a function of a parallel programming library optimized for a particular hardware platform. A result of executing the program is stored.

A method begins by generating application system state transitions from inputted requirements and parameters. For a current implementation of generating application code, the method continues by entering a loop. The loop begins by generating a current intermediate result based on a previous implementation and in accordance with current application code development factors. The loop continues by generating at least one test case based on the one or more of the application system state transitions. The loop continues by testing the current intermediate result in accordance with the at least one test case. When the testing is unfavorable, the loop continues by modifying one or more of: the one or more of the plurality of application system state transitions, the one or more of the parameters, and the one or more implementation tools. The loop then continues by repeating the loop using the modified current application code development factors.

Systems and methods decouple model components from a model execution style for which the model components are created, and the model components may be utilized in parent models having different execution styles. A model component may be partitioned into executable entities, and the entry points of the executable entities and their call styles may be identified. An adaptation layer that includes access points for the entry points may be constructed. The model component, including the adaptation layer, may be included in the model, and connection elements of the parent model may be connected to the access points of the adaptation layer. The execution call styles associated with the connection elements of the parent model may be bound to the execution call styles of the entry points as originally designed. The adaptation layer may manage translation of call styles and may coordinate scheduling of data communication with the model component.