Abstract: A flow graph of a first compilation unit of the program is expanded. The first corresponding parameter states of statements of the first compilation unit are marked as empty. The statements of the first compilation unit are processed to update the first corresponding parameter states with an exact type and a value set. Processing the statements include tracking a control flow predicate of a block in the first compilation unit, wherein the block comprises a first subset of the statements. Only after the control flow predicate indicates that the block is executable, the method includes processing an invocation statement in the first subset of the first statements, and expanding, responsive to the invocation statement, the flow graph for a second compilation unit of the program that is identified by the invocation statement.

Abstract: Techniques are disclosed for dynamically managing access to cross-cloud services. Provided are access control mechanisms for controlling and/or managing access to cross-cloud services offered by and between one or more cloud service providers. The techniques include detecting that a request for a cloud service has been received by a first component of a first cloud environment of a first cloud service provider and receiving an indication that deployment of the cloud service is permitted. In response to receiving the indication, a second component of the first cloud environment generates an instruction for implementing the cloud service within a second cloud environment and causes deployment of the cloud service within the second cloud environment based on the instruction.

Abstract: An architecture for offering a service of a first cloud service provider via a second cloud service provider is disclosed. A first cloud service provider infrastructure includes a first infrastructure and a second infrastructure. The first infrastructure is physically connected to a third infrastructure of a second cloud service provider infrastructure based on a first protocol. The first infrastructure is also physically connected to the second infrastructure based on a second protocol that is different from the first protocol. Using the first and second infrastructures, low latency high-bandwidth cross-cloud services can be provisioned and managed between private clouds of different cloud service providers.

Abstract: Techniques are disclosed for provisioning and managing resources within a cloud infrastructure of a first cloud service provider for cloud services offered by a second cloud service provider. Cross-cloud services can be provisioned and managed by and between private clouds of cloud service providers. The techniques include receiving a request for a cloud service by a component of a first private cloud within a first cloud environment and from a component of a second private cloud within a second cloud environment. The techniques further include the component of the first private cloud performing one or more operations to establish network connectivity prerequisites for network connectivity between the first private cloud and the second private cloud and causing one or more components of the first private cloud to provision the cloud service in the second private cloud.

Abstract: Techniques are described for partitioning B-tree keys of file systems into key ranges for parallel processing in delta generation during file storage replications between file systems in different cloud infrastructure regions. In certain embodiments, a delta generation processing for cross-region replication may utilize a key-range splitting mechanism involving a recursive algorithm that partitions B-tree keys of a source file system into roughly equal-size key ranges. All the partitioned key ranges may be processed in parallel and concurrently by different processing threads, one thread per key range, to improve the performance of the delta generation and achieve scalability.

Abstract: Techniques described herein relate to systems and methods of data storage, and more particularly to providing layering of file system functionality on an object interface. In certain embodiments, file system functionality may be layered on cloud object interfaces to provide cloud-based storage while allowing for functionality expected from a legacy applications. For instance, POSIX interfaces and semantics may be layered on cloud-based storage, while providing access to data in a manner consistent with file-based access with data organization in name hierarchies. Various embodiments also may provide for memory mapping of data so that memory map changes are reflected in persistent storage while ensuring consistency between memory map changes and writes. For example, by transforming a ZFS file system disk-based storage into ZFS cloud-based storage, the ZFS file system gains the elastic nature of cloud storage.

Abstract: Techniques are disclosed for provisioning a cloud service of a first cloud service provider using a control plane of a second cloud service provider. The techniques include detecting that a request for a cloud service provided by the first cloud service provider has been received from the second cloud environment of a second cloud service provider different from the first cloud service provider. The techniques further include, after detecting that the request for the cloud service has been received, provisioning a first set of resources within the first cloud environment and linking the first set of resources to a second set of resources within the second cloud environment. Linking the first set of resources to the second set of resources enables data pertaining to the cloud service to be transferred from the second cloud environment to the first cloud environment.

Abstract: Techniques for generating a retail forecasting model from product-cluster-based estimated elasticity values to forecast the effects of price changes on the demand for a set of products are disclosed. A system generates cluster-based price-elasticity values for a set of products by applying a set of regressive elasticity-estimation algorithms to a set of product data and clustering products based on product descriptions and estimated price-elasticity values. The system uses the cluster-based price-elasticity values for the products to generate the retail forecasting model.

Abstract: Techniques are disclosed for providing a scalable multi-tenant serve pool for chatbot systems. A query serving system (QSS) receives a request to serve a query for a skillbot. The QSS includes: (i) a plurality of deployments in a serving pool, and (ii) a plurality of deployments in a free pool. The QSS determines whether a first deployment from the plurality of deployments in the serving pool can serve the query based on an identifier of the skillbot. In response to determining that the first deployment cannot serve the query, the QSS selects a second deployment from the plurality of deployments in the free pool to be assigned to the skillbot, and loads a machine-learning model associated with the skillbot into the second deployment, wherein the machine-learning model is trained to serve the query for the skillbot. The query is served using the machine-learning model loaded into the second deployment.

Abstract: Techniques are disclosed for providing services based on infrastructure distributed between multiple cloud service providers. Low-latency high-bandwidth cross-cloud services can be provisioned and managed by and between private clouds of cloud service providers. The techniques include forming a cloud network between a first set of compute resources of a first infrastructure of a first cloud environment and a second set of compute resources of a second infrastructure of a second cloud environment. The first cloud environment is provided by a first cloud service provider and the second cloud environment is provided by a second cloud service provider different from the first cloud service provider.

Abstract: A method includes obtaining initial transcription for input natural speech; performing segmentation of initial transcription into text portions, based on punctuation marks in initial transcription; determining segment-level timestamps for text portions based on the input natural speech; performing audio segmentation on input natural speech, by cutting input natural speech based on segment-level timestamps, to obtain audio chunks; generating transcription portions for each of the audio chunks; merging transcription portions to form re-transcription; determining word-level timestamps for re-transcription, by aligning input natural speech against re-transcription; calculating silence time periods, each corresponding to silence between each two adjacent words of input natural speech, based on word-level timestamps; performing a final segmentation on input natural speech and re-transcription, based on silence time periods, to generate final audio segments and corresponding final transcription portions.

Abstract: Techniques are disclosed herein for transforming natural language conversations into a visual output. In one aspect, a computer-implement method includes generating an input string by concatenating a natural language utterance with a schema representation comprising a set of entities for visualization actions, generating, by a first encoder of a machine learning model, one or more embeddings of the input string, encoding, by a second encoder of the machine learning model, relations between elements in the schema representation and words in the natural language utterance based on the one or more embeddings, generating, by a grammar-based decoder of the machine learning model and based on the encoded relations and the one or more embeddings, an intermediate logical form that represents at least the query, the one or more visualization actions, or the combination thereof, and generating, based on the intermediate logical form, a command for a computing system.

Abstract: Techniques are disclosed for provisioning and managing resources within a cloud infrastructure of a first cloud service provider for a cloud service offered by a second cloud service provider. Cross-cloud services can be provisioned and managed by and between private clouds of cloud service providers. The techniques include receiving a request for a cloud service by a component of a first private cloud within a first cloud environment and from a component of a second private cloud within a second cloud environment. The techniques further include the component of the first private cloud performing one or more operations to establish network connectivity prerequisites for network connectivity between the first private cloud and the second private cloud and causing one or more components of the first private cloud to provision the cloud service in the second private cloud.

Abstract: Operations include determining whether to cache resources accessed by a query based on the execution time of the query. The system identifies a set of executions of a same query. The system determines a cumulative execution time for the set of executions of the same query. If the cumulative execution time exceeds a threshold value, then the system caches a resource used for execution of the query.

Abstract: The present disclosure relates to managing services by a managed service provider (MSP) in a cloud based infrastructure. A control plane of the MSP is established in a first tenancy, and a first access plane of the MSP is established in a second tenancy of a cloud environment. The control plane is configured to manage a plurality of services offered by the MSP to a first host machine included in the second tenancy. A first request is transmitted from the control plane to the first access plane, where the first request is forwarded by the first access plane to the first host machine, and corresponds to a service utilized by the first host machine and managed by the control plane of the MSP. In response to the first request being validated, a first state of the first host machine is modified in the second tenancy based on the first request.

Abstract: Various embodiments of the present technology generally relate to systems and methods for efficiently cleaning up resources within a network by bundling binding session audit requests. A binding support function (BSF) may determine support for bundled audit requests and unbundled audit requests at one or more policy control functions (PCFs) associated with potentially stale session binding records at the BSF. Based on audit support function at the PCF, the BSF may determine whether to bundle audit requests together, issue unbundled audit requests, or not issue any audit requests. PCF support for bundled or unbundled auditing may be provided via a network repository function (NRF), via headers in binding creation or update requests, or both.

Abstract: The present disclosure relates to chatbot systems, and more particularly, to batching techniques for handling unbalanced training data when training a model such that bias is removed from the trained machine learning model when performing inference. In an embodiment, a plurality of raw utterances is obtained. A bias eliminating distribution is determined and a subset of the plurality of raw utterances is batched according to the bias-reducing distribution. The resulting unbiased training data may be input into a prediction model for training the prediction model. The trained prediction model may be obtained and utilized to predict unbiased results from new inputs received by the trained prediction model.

Abstract: The present embodiments relate to a keystore service for encryption for a computing device of a cloud computing system. The computing device of the cloud computing system can receive a key identification, an encrypted key encryption key, an identity of a client device, and a request from the client device to store the key identification and the encrypted key identification key. The computing device can verify, based at least in part on the identity, an authority of the client device to store the key identification and encrypted key encryption key. The computing device can transmit, based at least in part on the verification, the key identification, the encrypted key encryption key, and request to store the key identification and the encrypted key encryption key to a keystore.

Abstract: Embodiments include systems and methods for generating a data throughput estimation model. A system may be monitored to measure both (a) data throughput and (b) computing statistics of one or more computing resources to generate an initial data set. The relationship between the data throughput and the computing statistics, in the initial data set, is used to generate a data throughput estimation model. The data throughput estimation model may be generated using a machine learning model, a neural network algorithm, boosting decision tree algorithm, and/or a random forest decision tree algorithm. Additional measurements of the computing resource statistics may be applied to the data throughput estimation model to estimate data throughput.

Abstract: Techniques for leveraging frequent patterns identified in a captured workload are provided. In one approach, multiple frequent patterns detected in a captured workload may be ordered by frequency to determine, for example, which patterns should be targeted for optimization. In another approach, a model of a captured workload is created, where the model comprises nodes that represent templates (which in turn correspond to requests) and edges that represent transitions between templates. The model is used to create an artificial workload, such as a workload that is twice as large as the originally-captured workload. The model may also be edited before creating the artificial workload. In another approach, workload models are compared to identify errors, regressions, or security issues. In another approach, an artificial workload is created for an application that is not yet deployed and then executed to determine whether the artificial workload or the originally-captured workload executed faster.