Abstract: A computer-implemented method includes receiving an event published by an application running on a server in response to a request for a service; filtering a full list of users of the server, using a tenant filter, to identify a group of tenant users who belong to a tenant identified by the event; filtering the group of tenant users, using a service filter, to identify a group of service-specific users interacting with the service identified by the event; filtering the group of service-specific users, using one or more user filters, to identify a group of eligible users permitted to receive the event; filtering the group of eligible users, using one or more topic filters, to identify a group of target users who subscribe to a topic of the event; and routing the event, via WebSocket connections, to a group of target clients associated with the group of target users.

Abstract: In an example embodiment, an iterative process is used to calculate the number of parallel work processes to set for a computer system. Specifically, an execution unit is started and the execution time for that execution unit is measured. This execution time for the single execution unit is called “unit time.” Then a fixed number (e.g., 20) of execution units are started, and the execution times of each are measured. If the execution time consumption of each of the execution units is lower than some fixed threshold percentage of the unit time (e.g., 120%), then this means that the maximum parallel processing capacity is higher than the fixed number of execution unit. Then more execution units can be added and the process repeated until the execution units' execution times exceed that fixed threshold percentage.

Abstract: Techniques for automatically revoking leaked access credentials are disclosed. In some embodiments, a computer system may receive an indication that a credential for accessing a resource has been leaked, where the credential has been leaked by being included in content that has been published on an online service or has been stored in a shared folder of the online service. The computer system may then determine that the credential is effective in accessing the resource, and, in response to the determining that the credential is effective, trigger a revocation of the credential, the revocation of the credential causing the credential to no longer be effective in accessing the resource.

Abstract: An enterprise Retrieval-Augmented Generation (“RAG”) system may be associated with at least one Large Language Model (“LLM”). A test dataset may include questions, context information, and ground truth answers. An enterprise RAG evaluation framework, coupled to the test dataset and the enterprise RAG system, iteratively perform the following process until an evaluation threshold is satisfied: determine an RAG generation metric score, determine an RAG retrieval metric score, determine an end-to-end RAG metric score, and automatically adjust at least one RAG turning parameter. When the evaluation threshold is satisfied, the framework may output a metrics report including the RAG generation metrics scores, the RAG retrieval metrics scores, and end-to-end RAG metrics scores.

Abstract: Example methods and systems relate to the modeling of production processes to include intermediate products. An enterprise resource planning (ERP) system receives user input to define an intermediate product, and generates a master data record for the intermediate product. The ERP system further receives user input to include the intermediate product in a production model within the ERP system. The production model is associated with a real-world production process. The production model is updated based on the master data record to integrate the intermediate product into a sequence of activities. The sequence of activities comprises a first activity that produces the intermediate product and a second activity that consumes the intermediate product. The ERP system executes a production order based on the production model. Execution of the production order includes tracking state information of the intermediate product within the ERP system.

Abstract: In an example embodiment, an LLM is used to identify similar microservices to allow a microservice that is similar to a dependent microservice that is down to be used instead of the downed dependent microservice until the downed dependent microservice can be brought back online. Specifically, the LLM is utilized in two different manners. First, it is used to generate an embedding for an API of each of multiple microservices in a system. These embeddings may then be used to retrieve similar APIs to the API of a downed microservice. Then the LLM can be further used to select the most qualified of the similar APIs, based on similar functionality and input/output parameters. The most qualified of the APIs can then be tested for final selection of a (temporary) replacement API that can be used in lieu of the API of the downed microservice.

Abstract: In an example embodiment, rather than use large language model (LLM) to directly generate desired computer code, an intermediate representation is generated by the LLM. The LLM is used to generate the portion of the computer code that cannot be computed programmatically (which may be called the “creative” part for purposes of the present disclosure). The intermediate representation can then be fed into a separate programmatic component that compiles the intermediate representation into compilable computer code. This fine-tuning may involve, for example, sanitizing the intermediate representation, enhancing the intermediate representation, and formatting the intermediate file, as well as modifying the intermediate representation based on a feature set.

Abstract: Embodiments of the present disclosure include techniques for recovering data. In one embodiment, data is copied to a buffer. A plurality of processing functions receive the data in the buffer as data pages and perform processing operations. The processed data pages are then stored in persistent memory. The main memory is monitored so that the main memory of the database is maintained in a state such that a consistent flow of data may be written to persistent memory during the recovery process.

Abstract: A method comprises generating vectors of sentences, the sentences including text, performing an augmentation operation on a first vector of the vectors and a second vector of the vectors, the augmentation operation comprising: merging the first vector with the second vector such that at least a first element of the first vector is combined with a second element of the second vector, wherein the first element is representative of a first sentence of the plurality of sentences and the second element represents a second sentence of the plurality of sentences, inputting, the first vector that is merged with the second vector, into a first layer, and generating a third vector that is based on the first vector that is merged with the second vector, the third vector including a third element that represents the first sentence and the second sentence.

Abstract: A system associated with a public key infrastructure certificate framework in a cloud computing environment may include a certificate authority data store that contains information about a plurality of certificate authority instances (with each certificate authority instance being associated with an instance index and an instance deployment time). A certificate authority server, coupled to the certificate authority data store, may retrieve an instance index and instance deployment time from the certificate authority data store. The certificate authority server may then determine a current certificate identifier generation timestamp. A unique certificate identifier for a public key certificate is generated by the certificate authority server based on a deterministic creation algorithm, the instance index, the instance deployment time, and the certificate identifier generation timestamp. The public key certificate can then be issued using the unique certificate identifier.

Abstract: Embodiments may be associated with a data source and a data service tool. A performance optimizer may determine a new type of data job to be executed based on a job execution parameter, perform a first execution of the new type of data job (such that data operations are performed at the data service tool), and collect first performance results. The performance optimizer then performs a second execution of the new type of data job (such that data operations are pushed down and performed at the data source) and collects second performance results. The first and second performance results are compared, and a result storage is updated with an indication of whether subsequent executions of the new type of data job will perform data operations at the data service tool or at the data source. The indication stored in the result storage may comprise, for example, a pushdown flag.

Abstract: A system and method include identification of a future fault period and a future fault amount associated with a component, determination of production characteristics of the component, generation of a multi-dimensional vector representing the future fault period, the future fault amount and the production characteristics, searching of a vector database of multi-dimensional vectors to identify a second multi-dimensional vector similar to the first multi-dimensional vector and first actions corresponding to the second multi-dimensional vector, and automatic execution of the first actions to resolve the future fault period and the future fault amount associated with the component.

Abstract: A system and method including generating a representation of an object associated with an equipment in an equipment management network, including associating a first identifier for the object in the equipment management network and specifying a first service integrated with the equipment management network to associate with the object; automatically transmitting an event notification from the equipment management network to the first service to inform the first service of the object in the equipment management network; receiving, from the first service, an indication of an external identifier associated with a replication of the object in the first service; correlating the first identifier associated with the object with the external identifier associated with the replication of the object in the first service; and storing a record of the correlation in a data store of the equipment management network.

Abstract: The present application is directed to a system which offers a turnkey solution (BPaaS) for running an organization business process via a third-party provider. In one example, the method may include generating a first plurality of finite state automata corresponding to a plurality of event sequences of a software process of a user, obtaining a second plurality of finite state automata based on a second plurality of event sequences of a host process of a provider, establishing channels of communication between the first and second plurality of finite state automata, executing the first and second plurality of finite state automata in parallel via the established channels, and determining whether the user process is compatible with the host process of the provider based on results of the execution, and displaying a result of the determination via a user interface of a computing system associated with the user.

Abstract: Various embodiments for a continuous learning system are described herein. An embodiment operates by receiving a query from a user and identifying an unknown phrase in the query. User feedback regarding the unknown phrase is requested and received. A first pre-existing entity of a plurality of pre-existing entities that corresponds to the received user feedback is identified. A relationship between the first pre-existing entity and the unknown is added to the knowledgebase. The query is executed against the knowledgebase using the first pre-existing entity. A response, to the executed query, is provided to the user.

Abstract: According to some embodiments, systems and methods are provided including a memory storing processor-executable program code; and a processing unit to execute the processor-executable program code to cause the system to: select an element from a plurality of elements, wherein each element includes one or more configurable parameters; configure each parameter for the selected element; apply the configured element to a task in a task list; and display the parameters on a user interface display based on the applied configured element. Numerous other aspects are provided.

Abstract: Disclosed herein are system, method, and computer program product embodiments for selectively anonymizing user data. An embodiment operates by receiving the user data, wherein the user data comprises first data and a first prompt, and the first prompt indicates how to process the first data. The embodiment then receives an anonymization template, wherein the anonymization template specifies a profile to be anonymized in the user data and a tool used for anonymization. The embodiment then creates, based on the anonymization template, anonymized user data by anonymizing the profile in the user data using the tool specified in the anonymization template. The embodiment then input the anonymized user data to a first language model. The embodiment then receiving an anonymized response, wherein the anonymized response is a result of the first language model processing the anonymized user data.

Abstract: Systems and methods include input of a description of a procedure to a large language model to determine a domain of the description, determination of modifiers to the description based on the domain, determination of example procedure models and descriptions corresponding to the example procedure models based on the domain, generation of a procedure model prompt based on the description, the modifiers, the example procedure models and corresponding descriptions, provision of the procedure model prompt to the large language model, reception, in response to the prompt, of a generated procedure model from the large language model, and storage of the generated procedure model for execution by a workflow automation system.

Abstract: Before performing dynamic parameter replacement, an object and a parameter-property mapping are obtained. The object may be an infrastructure template for provisioning physical or virtual computer systems. The object includes a plurality of properties and the plurality of properties include a plurality of placeholders. Each particular placeholder of the plurality of placeholders is a portion of a particular value of a particular property. The parameter-property mapping includes a plurality of key-value pairs where each key value pair includes a property key and a property value. To perform dynamic parameter replacement, the plurality of properties of the object (e.g., infrastructure template) are looped over using generic classes to identify and replace placeholders of the plurality of placeholders which match the property key with the property value.

Abstract: Techniques are disclosed for extracting and applying structured debugging knowledge to resolve software bugs. A computing system processes electronic representations of software bug reports, including comments describing reasoning processes for debugging. A first neural language model extracts structured reasoning paths from the reports, which are used to fine-tune a second neural language model. When a new bug is encountered, the second model generates suggested actions or deploys a fix. Alternatively, structured reasoning paths are embedded into a vector database, enabling similarity searches against new bug descriptions. Retrieved reasoning paths are submitted to a neural language model to generate debugging recommendations or automated fixes. A hierarchical structure preserves relationships between reasoning paths, facilitating reasoning path retrieval and refinement.