Abstract: A plurality of metrics records, including some records indicating metrics for which anomaly analysis has been performed, is obtained. Using a training data set which includes the metrics records, a machine learning model is trained to predict an anomaly analysis relevance score for an input record which indicates a metric name. Collection of a particular metric of an application is initiated based at least in part on an anomaly analysis relevance score obtained for the particular metric using a trained version of the model.

Abstract: Systems, devices, and methods are provided for end-to-end tagging and tracking of event signals subject to privacy policies. User events comprising event source data may be collected. A user may be associated with a segment, and the segment may be mapped to data lineage information corresponding to the events that were used to determine that the user is in segment. A segment cache may be generated comprising portions that correspond to different data lineages subject to different privacy policies. Bid requests may be received and processed according to applicable privacy policies.

Abstract: Methods, systems, and computer-readable media for dynamic prefetching for database queries are disclosed. A query of a database is started according to a first prefetch policy. Before completing the query, the first prefetch policy is changed to a second prefetch policy. A portion of the query is performed according to the second prefetch policy.

Abstract: A hierarchical embedding model is used to obtain respective language-agnostic embeddings of entity records of a cross-language data set. A plurality of record representation pairs is prepared based at least in part on the language-agnostic embeddings. A machine learning model is trained using the record representations pairs to generate similarity scores for pairs of entity records whose text attributes are expressed in different languages.

Abstract: Computer systems and associated methods are disclosed to implement the non-interactive join of privacy-preserving dataset sketches. In some embodiments, an entity can publish a one-time sketch of their dataset that would enable another entity to join their data without exposing private information. The sketch can map, using a hash function, the identities associated with a first value of the dataset to a data structure, in some embodiments. A same or different entity can join the first sketch with a privacy-preserving second sketch of a second dataset that includes added noise, and can determine an estimate of a number of identities that correspond with specific values of the first and second datasets from the joined dataset. The sketch can be published just one time, and therefore does not require separate new private computations with privacy budgeting for each additional party when a join is desired, in some embodiments.

Abstract: An application debugger may define multiple unique initial states of execution for an application on behalf of a client. These unique initial states may include unique sets of application variable values. The application debugger may then create respective application runtimes initialized with the respective sets of application variable values. Then, the execution runtimes may be independently executed to a termination condition that may include a defined set of breakpoints, a normal termination or an abnormal termination. The application debugger may then receive termination state information from the respective execution runtimes, provide additional analysis of these termination states and report results to the client.

Abstract: Methods for replicating transactional tables of a transactional database to an analytical database and maintaining updates to those transactional table representations are disclosed. Snapshots of the transactional tables are provided to the analytical database via a transport mechanism, such as a data storage service or a data streaming service, and stored at the analytical database. Then, checkpoints comprising portions of a change-data-capture log that has recorded transactional changes to the transactional tables of the transactional database are provided to the analytical database via the same or different transport mechanism and used to commit those transactional changes to the snapshot representations. The snapshot representations may be used to respond to incoming analytical queries in order to provide real-time querying results.

Abstract: A virtualized computing service provides a computing instance capable of requesting attestation of the authenticity of the hypervisor implementing the computing instances. An attestation device included in a virtualization host maintains a log of hash values representing hypervisor versions that have been implemented at the virtualization host. Also, an independent auditor (e.g., attestation service) is provided software configurations that are known to be authentic. The independent auditor generates hash values for the authentic hypervisor versions. The computing instance receives a response from the local attestation device indicating hash values of hypervisor versions currently and/or previously deployed on the virtualization host, and the computing instance forwards the hash values to the independent auditor to authenticate that they match the hash values of the known authentic hypervisor versions.

Abstract: Methods for implementing insert, update, and delete events of a change-data-capture log in transactional order to a representation of a transactional table are disclosed. Insert events may be applied and committed to the representation, while delete events may be buffered and committed to a shadow table. Such buffering may allow the delete events to be applied and committed as a batch to the given representation as a later background operation, allowing for an optimized application of the change-data-capture log to the representation. Once the delete events are committed to the representation, they are removed from the shadow table. Update events may be similarly treated in which the insert event component of the update event is applied and committed to the representation while the delete event component is committed to the shadow table, before being later committed to the representation and removed from the shadow table.

Abstract: Real-time low-complexity stereo speech enhancement with spatial cue preservation may be performed. A stereo speech enhancement system receives a stereo input signal (e.g., a left and right input signal). The stereo speech enhancement system estimates spatial cues for a target speaker and downmixes the stereo input signal into a monaural signal. A low-complexity model may then process the monaural signal to generate an enhanced monaural signal. The stereo speech enhancement system upmixes the enhanced monaural signal based on the estimated spatial cues for the target speaker, to generate an enhanced stereo output signal.

Abstract: Monitoring may be performed for time series prediction models. Data to generate a new time series forecast may be received. A determination may be made that the data is associated with a previously generated time series forecast by a machine learning model. Performance metrics may be generated for the machine learning model according to a comparison of the data with the previously generated time series forecast. The performance metrics can then be provided for further analysis and action.

Abstract: Automatic onboarding of heterogeneous edge devices onto a client network may be performed (e.g., different types of edge devices manufactured by different vendors). A device onboarding and integration service may determine a type of edge device to be onboarded and based on the type of edge device, determine a particular type of device attestation to perform. Upon verifying the device, the service establishes a secure connection and registers the device. The service configures an agent and sends the agent to the device. The service also sends, to the device, a recipe of different locations for the device to download software components and to subscribe to services in order to complete the onboarding process.

Abstract: Methods for negotiating parameters for maintaining, at an analytical database, a replica of a transactional table, stored at a transactional database, are disclosed. By establishing parameters of said maintenance prior to commencing, the transactional database and the analytical database may remain only loosely coupled after completing the negotiation. The parameters of the negotiation may be stored in a location that is accessible to both the transactional database and the analytical database, such that the results of the negotiation may be accessed during the process of maintaining the replica. The negotiation may include a procedure to provide portions of a change-data-capture log, associated to the transactional table, to the analytical database via a transport mechanism such that the replica may be updated, in addition to a list of agreed upon data definition language commands that may be used in the portions of the change-data-capture log.

Abstract: Evaluation data sets may be programmatically generated for code generation models. An evaluation data set is obtained that includes items that correspond to different evaluation tests for a code generation system. The individual items of the evaluation data set maybe converted, including the conversion of a function signature for the items, the test statements for the items and using a code generation system to generate the body of the function.

Abstract: A data source configured to provide a representation of an environment of one or more agents is identified. Using a data set obtained from the data source, a neural network-based reinforcement learning model with one or more attention layers is trained. Importance indicators generated by the attention layers are used to identify actions to be initiated by an agent. A trained version of the model is stored.

Abstract: Pre-fix matching may constrain the generation of next token predictions. Input text to perform a next token prediction may be received. Multiple tokens may be determined from the input text, including a partial token. From possible tokens, one or more matching possible tokens with the partial token may be identified. Next token predictions may then be filtered using the identified possible tokens in order to ensure that the partial token is matched.

Abstract: Random token segmentation may be implemented for next token prediction. Text data may be received for training a machine learning model to predict a next token given input text tokens. Multiple tokens may be determined from the text data. Different ones of the multiple token may be randomly segmented in to sub-tokens. The machine learning model may then be trained using the multiple tokens including the respective sub-tokens as a training data set.

Abstract: Code completion suggestions may be proactively obtained and validated. An event that triggers obtaining a code completion suggestion for inclusion in a code file being edited using an integrated development environment may be detected. The code completion suggestion may be obtained. The characters of the code completion suggestion may be compared with characters added to the code file after the detection of the event that triggered obtaining the code completion suggestion to determine whether the code completion suggestion is valid. A valid code completion suggestion may then be displayed.

Abstract: Compression profiles may be searched for trained neural networks. An iterative compression profile search may be performed response to a search request. Different prospective compression profiles may be generated for trained neural networks according to a search policy. Performance of compressed versions of the trained neural networks according to the compression profiles may be tracked. The search policy may be updated according to an evaluation of the performance of the compression profiles for the compressed versions of the trained neural networks using compression performance criteria. When a search criteria is satisfied, a result for the compression profile search may be provided.

Abstract: Instructions stored at a memory of an offloading card of a server, when executed at a processor of the offloading card, perform virtualization management tasks pertaining to a compute instance launched at the server, including allocation of memory of the server to the compute instance. In response to a request from the compute instance, a network function of a radio-based application is executed at an accelerator incorporated within the offloading card.

Abstract: A first set of instructions, which is provided access to a first address space, is scheduled for execution at a first hardware thread of a processor. Prior to executing an instruction of a second set of instructions, which accesses a second address space, at the first hardware thread, a determination is made that the second address space is accessible from a second hardware thread of the processor.

Abstract: A first security rule set for traffic of a compute instance run at a virtualization server is stored in response to a programmatic request. A second security rule set for traffic of a network function accelerator of the virtualization server is stored in response to another programmatic request. Prior to delivery of network messages to the compute instance or from the accelerator, compliance with the applicable security rule set is verified.

Abstract: Respective network addresses are assigned to a networking hardware device and a network function accelerator incorporated within an offloading card of a virtualization server by a control plane server of a virtualized computing service. A compute instance is launched at the virtualization server using a virtualization controller of the offloading card. The compute instance executes a network function of a radio-based application in response to a request received using the network address assigned to the hardware device, and requests execution of a second network function at the accelerator. The result of the second network function is sent to a radio unit of the application using the address assigned to the accelerator.

Abstract: Apparatus and methods are disclosed herein for remote, direct memory access (RDMA) technology that enables direct memory access from one host computer memory to another host computer memory over a physical or virtual computer network. In one example, RDMA functionality is provided by combining a host executing instructions for RDMA software applications with dedicated hardware accelerator, where the host has access to substantially all RDMA connection context (e.g., configuration, status, state machine, counters, etc.) and a hardware accelerator stores a portion of the context information for a set of currently active transactions. The hardware accelerator derives a transaction identifier from header information in received RDMA packets and performs a local RDMA operation using at least a portion of the received data, a destination address of the location RDMA operation being based at least in part on a memory address determined using the transaction identifier.

Abstract: Centralized data set management techniques are implemented for data sets for machine learning applications. New data sets may be created from a search of managed data set labels. Data objects of the managed data sets that are found in the search are then used to generate a new data set. The new data set is then made available for machine learning applications to use.

Abstract: Methods, systems, and computer-readable media for a cloud-based database for spatial data lifecycle management. A spatial database receives elements of spatial data from a plurality of clients of the distributed spatial database. An individual element of the spatial data comprises one or more location values or one or more spatial objects. The spatial database stores the plurality of elements of spatial data using a plurality of storage resources. The spatial database receives a query. The spatial database determines one or more elements of spatial data matching the query from the plurality of elements of spatial data that were stored using the plurality of storage resources.

Abstract: A system that implements a scalable data storage service may maintain tables in a data store on behalf of storage service clients. The service may maintain table data in multiple replicas of partitions that are stored on respective computing nodes in the system. In response to detecting an anomaly in the system, detecting a change in data volume on a partition or service request traffic directed to a partition, or receiving a service request from a client to split a partition, the data storage service may create additional copies of a partition replica using a physical copy mechanism. The data storage service may issue a split command defined in an API for the data store to divide the original and additional replicas into multiple replica groups, and to configure each replica group to maintain a respective portion of the table data that was stored in the partition before the split.

Abstract: Replicated instances in a database environment provide for automatic failover and recovery. A monitoring component can periodically communicate with a primary and a secondary replica for an instance, with each capable of residing in a separate data zone or geographic location to provide a level of reliability and availability. A database running on the primary instance can have information synchronously replicated to the secondary replica at a block level, such that the primary and secondary replicas are in sync. In the event that the monitoring component is not able to communicate with one of the replicas, the monitoring component can attempt to determine whether those replicas can communicate with each other, as well as whether the replicas have the same data generation version. Depending on the state information, the monitoring component can automatically perform a recovery operation, such as to failover to the secondary replica or perform secondary replica recovery.

Abstract: A service implemented at a first isolated virtual network of a provider network is added to a database of privately-accessible services. Configuration changes that enable network packets to flow between the first isolated virtual network and a second isolated virtual network without utilizing a network address accessible from the public Internet are implemented. Service requests originating at the second isolated virtual network are transmitted to the first isolated virtual network via private pathways of the provider network. Metrics corresponding to service requests directed from the second isolated network to the service are collected and provided to the respective owners of one or both isolated virtual networks.

Abstract: Techniques are described for managing access of executing programs to non-local block data storage. In some situations, a block data storage service uses multiple server storage systems to reliably store network-accessible block data storage volumes that may be used by programs executing on other physical computing systems. A group of multiple server block data storage systems that store block data volumes may in some situations be co-located at a data center, and programs that use volumes stored there may execute on other physical computing systems at that data center. If a program using a volume becomes unavailable, another program (e.g., another copy of the same program) may in some situations obtain access to and continue to use the same volume, such as in an automatic manner in some such situations.

Abstract: A spatial search may be performed using representations of geometric shapes stored in a key-value store. A request to perform a spatial search may be received, the request including a geometric shape composed of one or more points. The points of the geometrical shape may be translated into one or more spatial indexes representing spatial cells using a space-filling curve. A key-value store may then be incrementally searched for each spatial index to identify spatial cells intersecting the geometric shape for which other known geometric shapes exist. The key-value store may then be searched to identify the known geometric shapes intersecting the geometric shape included in the search.

Abstract: At an artificial intelligence system, during a labeling feedback session, a visualization data set is presented via a programmatic interface. The visualization data set comprises a representation of data items for which labeling feedback is requested for generating a training set of a classifier. At least one of the data items is selected based on an estimated rank with respect to a metric associated with including the data item in a training set. During the session, respective labels for the data items and a filter criterion to be used to select additional data items are obtained. A classifier trained using the labels is stored.

Abstract: A storage service supports attachment of multiple clients to a distributed storage object and further supports persistent reservations that govern types of access the respective clients are granted with respect to the distributed storage object. In order to efficiently distribute reservation state changes to multiple partitions of the distributed storage object hosted by different data storage units/servers, existing connections are used between the data storage units/servers hosting the partitions of the distributed storage object and the connected clients to propagate reservation state changes.

Abstract: Compiled portions of code generated to perform a query plan at a query engine may be shared with other query engines. A data store, separate from the query engines, may store compiled portions of query code generated for different queries. If a query engine does not have a locally stored compiled portion of query code, then the separate data store may be accessed in order to obtain a compiled portion of query code, allowing reuse of compiled query code across different queries engines for queries directed to different databases.

Abstract: Code may be dynamically routed to computing resources for execution. Code may be received for execution on behalf of a client. Execution criteria for the code may be determined and computing resources that satisfy the execution criteria may be identified. The identified computing resources may then be procured for executing the code and then the code may be routed to the procured computing resources for execution. Permissions or authorization to execute the code may be shared to ensure that computing resources executing the code have the same permissions or authorization when executing the code.

Abstract: A vehicle information service implemented on one or more computers of a service provider network implements a first application programmatic interface (API) that allows a client to define inclusion parameters and a sample size for a fleet of vehicles from which vehicle data is to be collected. The vehicle information service also implements a second API that notifies the client when the requested vehicle data has been collected from the vehicle fleet. Additionally, the vehicle information service provides the client access to the collected vehicle data. The vehicle information service manages the collection of the vehicle data from the client defined vehicle fleet without requiring further client involvement and notifies the client when the collection of the vehicle data is complete.

Abstract: Methods and apparatus for partitioned private interconnects to provider networks are described. At least a portion of available bandwidth of a private physical interconnect between a provider network and a connectivity intermediary's network is designated as the bandwidth limit of an interconnect partition set up on behalf of a customer at the request of the intermediary. The intermediary's network comprises one or more devices to which at least one of the customer's devices is connected. Access to one or more resources of the provider network via the interconnect is enabled. Traffic monitoring results associated with the interconnect are used to enforce the designated bandwidth limit of the partition.

Abstract: A system for applying metadata tags based on metadata tag groups is described. The system includes a plurality of compute nodes configured to provide a plurality of computing resources for clients of a provider network and a resource tag manager for the provider network. The resource tag manager is configured to receive, from a first client of an account of the provider network, a creation request to establish a resource attribute group comprising a plurality of resource keys and a plurality of resource values, wherein respective metadata keys correspond to respective resource values. The resource tag manager may receive, from a second client of the account, an application request to apply the resource attribute group to a computing resource and apply the plurality of resource keys and the plurality of resource values to metadata maintained for the computing resource.

Abstract: A network-based data store maintaining storage volumes for file systems at storage clients may implement dynamic storage volume scaling. A scaling event may be detected according to a specified scaling policy for a storage volume maintained at a network-based data store that is mounted at a storage client and configured for a file system. In a least some embodiments, the scaling policy for the storage volume may be received from the storage client. In response to detecting the storage event, storage maintaining the storage volume may be modified according to the scaling policy to scale the storage volume. An indication of the scaling event may be sent to the storage client in order to update the file system at the storage client so that the file system may utilize the scaled storage volume.

Abstract: A first analysis based on at least a subset of collected health metrics of a monitored resource is performed at a health monitoring service. If the first analysis indicates that the probability that the monitored resource is in an unhealthy state is above a threshold, a mitigation action that includes directing at least some of the workload away from the monitored resource is implemented. In addition, a remediation analysis of the monitored resource is started to determine, using at least one additional health metric that was not examined in the first analysis, whether a longer-term remediation action is to be initiated.

Abstract: Inferences may be obtained to handle access requests at a non-relational database system. An access request may be received at a non-relational database system. The non-relational database system may determine that the access request uses a machine learning model to complete the access request. The non-relational database system may cause an inference to be generated using data items for the access request as input to the machine learning model. The access request may be completed using the generated inference.

Abstract: Connectivity between a radio-based application pipeline processing server and a control plane of a provider network is verified. Based on requests received at the control plane, a first isolated request handler, a second isolated request handler and an offloading manager are launched at the server. The offloading manager causes a first network function for which a request is received from the first request handler to be executed at a first network function accelerator of the server, and a second network function for which a request is received from the second request handler to be executed at a second network function accelerator of the server.

Abstract: Interactive assistances for executing natural language queries to data sets may be performed. A natural language query may be received. Candidate entity linkages may be determined between an entity recognized in the natural language query and columns in data sets. The candidate linkages may be ranked according to confidence scores which may be evaluated to detect ambiguity for an entity linkage. Candidate entity linkages may be provided to a user via an interface to select an entity linkage to use as part of completing the natural language query.

Abstract: A system and method for providing quantum entanglement using a hybrid space-fiber quantum network are described. The hybrid space-fiber quantum network includes a communications hub located proximate to an optical ground station and also includes an aerial entangled particle source, such as an entangled photon source attached to a satellite, drone, aircraft, etc. An atmospheric or free-space channel is used to distribute quantum entanglement between optical ground stations that are separated by geographic distances, via the aerial entangled particle source. Also, fiber optic links are connected to the communications hub located proximate to the optical ground station. The communications hub includes optical switches that enable any of the fiber optic links connected to the communications hub to receive or send distributed quantum entanglement to a remotely located recipient endpoint via the atmospheric or free-space channel.

Abstract: A processing service of a provider network may protect media content from being tampered with when it is transmitted from the provider network/transcoder to untrusted networks (e.g., third-party networks/CDNs) and to a media player. The processing service (e.g., the transcoder) generates a public and a private key. The service uses the private key to digitally sign content portions (e.g., video frames) before distribution to untrusted CDNs. The provider network creates a manifest that includes the public key. To play the media content, the media player obtains a manifest that includes the public key (via a secure/trusted connection with the provider network). The media player may then obtain the media content from an untrusted edge server/CDN and validate it using the public key that was separately obtained from the manifest (to verify the content was not tampered with).

Abstract: A quantum computing service provides a quantum algorithm development kit that enables a customer to define a quantum task, a quantum algorithm, or a quantum circuit using an intermediate representation. The quantum computing service is then configured to automatically translate the quantum task, quantum algorithm, or quantum circuit into a specific representation specific to a particular quantum computing technology selected by the customer to be used to execute the customer’s quantum task, quantum algorithm, or quantum circuit.

Abstract: Code search is used to support various features. Code segments may be indexed using a code structure representation of the code segment. Code segments may be compared for similarity to identify a match with an input code segment using a comparison of logic trees generated for the input code segment and a stored code segment in an entry with a matching index value.

Abstract: Systems and methods are disclosed for implementing a vehicle shadow service that includes a vehicle and/or vehicle shadow update mechanism. A vehicle shadow management sub-system of the vehicle shadow service includes or interfaces with an interface configured to receive vehicle state change requests. Received requests are stored in an update queue for a vehicle shadow and a validation processor determines whether the received state change requests are valid before sending a given state change request to a vehicle corresponding to the vehicle shadow. In some embodiments, state change requests may be invalidated due to being rendered moot by subsequently received requests, changes in state of the vehicle, expiration, etc. The corresponding vehicle shadow is then updated when a data stream from the vehicle indicates that the requested state change has in fact been implemented at the vehicle.

Abstract: Systems, devices, and methods are provided for facet-based context-aware user interfaces. A system such as an online purchasing system may determine one or more intent contexts based at least in part on one or more user interactions with an online purchasing system associated with a first product. A list of substitute products may be determined based on the intent context and facets associated with the substitute products to provide for a diversity of dimensions along which a user can explore different products.

Abstract: Systems and methods are disclosed for implementing perspective-based vehicle shadows. A user of a vehicle shadow service can specify different vehicle shadows for a given vehicle, wherein the different vehicle shadows comprise at least partially different sets of sensor data received from the vehicle and represent the vehicle from different perspectives. For example, a first vehicle shadow may represent the vehicle from the perspective of a first vehicle system, such as the tires, and another vehicle shadow may represent the same vehicle from the perspective of a second vehicle system, such as the engine. Streaming data is provided to the vehicle shadow service and is mapped to the respective perspective-based vehicle shadows, based on user defined configurations for the respective perspective-based vehicle shadows.