Abstract: Disclosed are some implementations of systems, apparatus, methods and computer program products for processing consent information. A data consolidation system obtains, from each of two or more computing systems, a corresponding set of data objects such that two or more sets of objects are obtained. Each data object is associated with a corresponding individual and includes consent information indicating whether the individual consents to an action in relation to corresponding contact information of the individual. Consent information is extracted, for at least one individual, from the sets of data objects such that two or more sets of consent information are collected, the sets of consent information including a first set of consent information corresponding to a first computing system and a second set of consent information corresponding to a second computing system.

Abstract: Disclosed herein are system, method, and computer program product embodiments for implementing variable a declarative authentication engine. The system generates a schema that includes a field and has a format defined by an authentication protocol associated with a service. The system then validates a connection request based on comparing the field of the generated schema to a field of the connection request for the service, wherein the connection request is formatted according to the schema and received from a client device. The system then provides the client device access to the service according to the connection request based on a result of the validating.

Abstract: Disclosed herein are system, method, and computer program product embodiments for implementing grounding of prompt templates using record snapshot. An embodiment operates by parsing a data record to generate a hierarchical tree graph of context data corresponding to a prompt template selected based on a type of a prompt request, and the data record is identified based on a data-record reference in the prompt request. The embodiment then generates a prompt from the prompt template by grounding the prompt template with a data object comprising the hierarchical tree graph of context data. The embodiment then queries a large language model with the prompt to generate an output specific to the data record.

Abstract: Techniques are disclosed relating to a database recovery routine to start up a database system in response to a database failure. The database system accesses checkpoint information identifying a set of active database transactions that were active at a flush point that occurred before the database failure. As a part of the database recovery routine, the database system replays database transactions that occurred between a recovery point and the flush point. The database transactions include the set of active database transactions but exclude any committed or aborted database transactions that occurred between the recovery point and the flush point such that less than a total number of database transactions occurring between the recovery point and the flush point are replayed. The database system further replays, without excluding committed or aborted database transactions, database transactions occurring between the flush point and a recovery end point at which the database failure occurred.

Abstract: Embodiments described herein provide systems and methods for training neural network based language models using human feedback. An existing (or generated) summary of a document is provided, and that summary may be used to generate a number of other summaries. A human annotator may reject the summary if there is any factuality issue with the summary. Summaries which are agreed to have no factuality problems are used as baseline summaries. Small atomic edits are made to the baseline summaries (e.g., replacing a single word or phrase) to create a group of summaries. Human annotators label each of these summaries as factual or not. The annotated summaries are used to train a summarization model and/or a factual detector model.

Abstract: A method receives audio data from a call. Services are performed to process the audio data to automatically generate a response, wherein the services include converting the audio data to input text, inputting the input text into a model to automatically generate a text response, and converting the text response to an audio response. Supplemental words are selected based on the input text. The method determines a type of service based on services performed to generate the audio response and determines a position in the response to insert the supplemental words based on the type of service. The supplemental words are provided for insertion in the call at the position to supplement the audio response.

Abstract: A method of data processing is described. The method includes converting a plain-text query into a vector-based object by using a text embedding function to process one or more tokens in the plain-text query. The method further includes retrieving a set of passages from a first datastore of the data processing system based on using one or more search indexes stored in a second datastore of the data processing system to compare the vector-based object and the one or more tokens in the plain-text query to vector-based objects and token-based objects associated with the set of passages. The method further includes generating a prompt that includes tokens from the plain-text query, tokens from one or more of the set of passages retrieved from the first datastore, and instructions for creating a response to the plain-text query. The method further includes transmitting the prompt to a large language model (LLM).

Abstract: Techniques for generating a virtual space overview are discussed herein. A communication platform may receive a request from a user profile to view an overview of a virtual space. Based on the request, the communication platform may receive criteria associated with the request. The criteria may indicate a type of content to include in the overview, a recency of the content, filters, etc. The communication platform may identify the content to include in the overview based on the criteria. The communication platform may use the content to generate the overview. That is, the overview may include one or more sections that include summaries and/or relevant data based on the content. The communication platform may cause overview to be displayed via a user interface of a user device of the user profile.

Abstract: A recommendation service access a data matrix listing of products associated with product profiles, the data matrix having product entries that store sparse historical electronic activity. For a target product it is determined which other products should be used to boost the historical electronic activity of the target product based on a first subset of product profiles that share product characteristics with the target product. Similarity scores are computed between the product profile of the target product and the first subset of product profiles to identify a second subset of one or more products having a similarity score above a scoring threshold. The historical electronic activity of the target product is boosted using the historical electronic activity of the other products in the second subset. Association values are calculated between the target product and the other products in the second subset by based on the boosted activity.

Abstract: A computer system permits installation of software by an independent software vendor (ISV) in different manners, such as within a user space of an organization tenant of the computer system, a space of the ISV within the computer system, or a space of the computer system itself. Packaging and installation functionality of the computer system performs necessary configurations so that an ISV can simply specify how its code is to be made available to users via the computer system, without itself needing to create different configurations depending on the desired manner of usage.

Abstract: A method for providing solutions to a user request, the method comprising: processing a user request identifying an assigned topic, wherein the user request is received from a remote graphical user interface (GUI) coupled to the processor, determining, by an updater coupled to the processor, a plurality of solutions to address the user request with the assigned topic, wherein the plurality of solutions are also determined based on recurring patterns associated with the user request with the assigned topic, transmitting, by the processor, the plurality of solutions received from the updater to the GUI, and identifying, by the processor, a resolution effectiveness of the plurality of solutions applied on the GUI.

Abstract: System, method and interface for generating data visualizations are provided. The system receives a user input to specify a natural language command directed to a data source. The system also generates a prompt for generating a data visualization based on relevant data fields and data values, rules that characterize the data visualization, and a context free grammar. The system also prompts a trained large language model using the prompt to generate a structured document following a domain-specific schema based on a shorthand notation. The system also uses a parser that uses the context free grammar to map the structured document to a visual specification. The visual specification specifies the data source, visual variables, and data fields from the data source. The system also generates and displaying a data visualization based on the visual specification, including displaying visual marks representing data, retrieved from the data source, for the data fields.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for backing up a data object in blocks. One of the methods includes determining, for a data object of a backup process, whether a size of the data object or an estimated backup time of the data object satisfies a criterion that, when satisfied, indicates that at least two blocks of the data object should be separately fetched from the source system by different workers; determining one or more markers for end points of the at least two blocks using data from a prior backup of the data object; and causing, at least partially concurrently for two or more blocks from the at least two blocks, a respective backup worker to fetch the respective block from a source system using at least one marker from the one or more markers that defines an end of the respective block.

Abstract: Embodiments described herein provide label modular prompts for a text classification task. A label modular prompt generator may determine a set of class labels of interest from a set of possible class labels associated with an input text sequence. The label modular prompt generator may generate a plurality of label prompts based on the set of class labels of interest. A first class label and a sequence of soft tokens that are generated based on representations associated with the first class label are concatenated into a first label prompt. The soft tokens are tunable using a plurality of parameters of the label modular prompt generator. The label modular prompt generator may provide an input of the input text sequence prepended with the plurality of label prompts to a pretrained language model. The pretrained language model may generate a task output in response to the input text sequence.

Abstract: A database system may include one or more relational databases storing information for a plurality of tenants in accordance with database object definitions. The database system may also include a communication interface providing the plurality of tenants with access to web applications through which to access the information and configured to receive an indication of one or more of the database object definitions from a tenant. The database system may also include a storage device storing a prompt template specific to the tenant and that includes one or more natural language instructions for generating text and a reference to the one or more database object definitions. The database system may also include a processor configured to retrieve a database record associated with the tenant and corresponding to the one or more database object definitions and to determine a text generation prompt based on the database record and the prompt template.

Abstract: Embodiments described herein provide a method for configuring an artificial intelligence (AI) conversation bot to respond to a user query based on retrieved contextual documents. The method includes: receiving, via a communication interface, a user query comprising a natural language description of a topic; generating, by a first neural network based language model, one or more subtopics of the topic based on a first input prompt combining the topic and a first instruction to generate the one or more subtopics; generating, by the first neural network based language model, one or more statements for at least one of the subtopics based on a second input prompt combining the one or more subtopics and a second instruction to generate the one or more statements; and generating, by the first neural network based language model, at least one document containing a set of randomly selected statements from the one or more statements.

Abstract: Systems and methods are provided for receiving, at a server, a workflow definition and generating a unique key for the received workflow definition. A distributed log storage may store the internal workflow schema having the not-started states to a state topic of the distributed log storage using the generated unique key, where the state topic includes the states of the internal workflow schema. One or more workers at the server may perform at least one operation based on a received message. The state may be updated at the distributed log storage based on the performed at least one operation. The state topic of the internal workflow schema for the generated key may be compacted based on the updated state, where the compacting reduces the states of the internal workflow schema to the current states, without intermediary states.

Abstract: An embodiment is a method performed by one or more electronic device to manage deployment of a configuration object for a machine learning application in a multi-tenant on-demand model serving infrastructure. The method includes receiving an indication that a configuration object for the machine learning application is ready to be deployed, obtaining the configuration object for the machine learning application in response to receiving the indication, and causing a router component to instantiate a router instance for the machine learning application, wherein the router component instantiates the router instance for the machine learning application based on the configuration object for the machine learning application and the router instance for the machine learning application is configured to route incoming scoring requests associated with the machine learning application to a scoring container that provides scoring functionality for the machine learning application.

Abstract: Techniques are disclosed for processing record writes in a database system having a log owner and one or more log tailers. The log owner maintains a transaction log, and the log tailers replay the transaction log to recreate a state of the log owner. A log tailer may receive a write request to perform a record write operation to write a base record to a database as part of a database transaction. The log tailer may determine additional records to write as a result of having to write the base record. The log tailer buffers the base record and additional records in a memory buffer and issues them as a single batch to the log owner to update the transaction log. The log owner may update the transaction log and return a response to the log tailer that indicates whether the log owner successfully updated the transaction log.

Abstract: A computer services environment may include web servers providing access domains and a network ingress paths receiving application-layer request messages. The application-layer request messages may each be received from a respective source via a respective ingress path and may be directed to a domain. The computing services environment may also include an orchestration engine configured to determine and implement mitigation policies corresponding with the ingress paths based on a classification of a subset of the plurality of application-layer request messages as being sent from a subset of the sources associated with a distributed denial of service attack. The mitigation policies may include rules to prevent a subset of subsequent application-layer request messages from the subset of the sources from reaching one or more components of the computing services environment.

Abstract: Systems, devices, and techniques are disclosed for transaction coordination across services. A service including virtual threads receives a transaction that is part of a multi-part transaction. The service determines the transaction does not already exist in transaction state data. The service updates the status of the transaction in the transaction state data to pending. The service processes the transaction. The service updates the status of the transaction in the transaction state data to completed. The service uses receives at and endpoint a request for the status of the transaction. The service responds to the request for the status of the transaction. The service sends a request for status to an endpoint another service that is processing another transaction of the multi-part transaction. The service receives a response from the another service of a status of completed.

Abstract: Embodiments described herein provide A method for training a neural network based model. The methods include receiving a training dataset with a plurality of training samples, and those samples are encoded into representations in feature space. A positive sample is determined from the raining dataset based on a relationship between the given query and the positive sample in feature space. For a given query, a positive sample from the training dataset is selected based on a relationship between the given query and the positive sample in a feature space. One or more negative samples from the training dataset that are within a reconfigurable distance to the positive sample in the feature space are selected, and a loss is computed based on the positive sample and the one or more negative samples. The neural network is trained based on the loss.

Abstract: A computing system may process entity name variations. The entity name variations may be transformed into vector representations. The vector representations may be stored in a database. The entity name variations may be clustered into groups of similar entity name variations based on a similarity metric between the vector representations. A canonical identifier for each group of similar entity name variations may be stored in the database. Each canonical identifier identifying a distinct entity.

Abstract: A computing device displays first data describing a dataset. At least a portion of the first data is encoded with metadata that links the first data to data values and/or data fields of the dataset. The computing device receives a user interaction with a first affordance. The user interaction specifies a first portion of the first data, which includes at least a first data field of the dataset. In response to receiving the user interaction, the computing device retrieves metadata corresponding to the first portion of the first data, and generates second data describing the dataset according to (i) the at least the first data field and (ii) data values of the at least the first data field specified in the metadata, corresponding to the first portion of the first data. The computing device concurrently displays the first data and the second data describing the dataset.

Abstract: A method for data processing is described. The method includes receiving an indication of a reference string from a cloud client of a content generation service. The method further includes generating multiple candidate strings associated with the reference string based on using a machine learning model to calculate similarity metrics between the reference string and the multiple candidate strings, where the machine learning model is trained using a dataset of annotated strings. The method further includes selecting a quantity of the candidate strings based on filtering the multiple candidate strings according to the similarity metrics. The method further includes causing the quantity of candidate strings to be displayed at the cloud client. The method further includes receiving feedback associated with the quantity of candidate strings and a selection of at least one candidate string displayed at the cloud client.

Abstract: Systems, methods, and devices are disclosed herein that provide security for requests sent to services in service meshes. A computing platform may be implemented using a server system. The computing platform is configurable to cause receiving a request from a service in a cloud-based computing environment, and identifying a chain of trust embedded in a portion of the request, the chain of trust being generated by one or more security entities in the cloud-based computing environment, the chain of trust identifying results of one or more security verification operations performed on the request. The computing platform is further configurable to cause determining if the chain of trust is a valid chain of trust based, at least in part, on one or more security policies, and sending the request to another entity in the cloud-based computing environment in response to determining the chain of trust is a valid chain of trust.

Abstract: Database systems and methods are provided for managing relationships between records in a database. An exemplary method involves identifying a virtual container for a first record at the database system based on one or more fields of the first record and updating the virtual container to associate the first record with one or more additional records. In response to an operation with respect to at least one of the records, the method binds an instance variable of a virtual container based on a relationship between respective values associated with the respective records, automatically initiate an action at the database system with respect to the one or more respective records satisfying the relationship constraint, and update a graphical user interface (GUI) display at a client device over a network to include graphical indication of the action at the database system responsive to the bound value for the instance variable.

Abstract: Techniques are described for securing secrets in software build workflows. In some implementations, build instructions call for execution of a first program module and a second program module, where the first program module has been approved to make a privileged request, but the second program module has not. The first program module can be stored in a trusted repository, separately from the second program module. When the first program module is loaded for execution, a cryptographic signature can be validated to determine that the first program module is authentic and as a condition for passing a privileged credential to the first program module. The second program module has no access to the privileged credential. Instead, when the second program module is loaded for execution, a determination can be made whether the second program module makes any privileged requests. Any privileged requests from the second program module will not be fulfilled.

Abstract: Methods, storage systems and computer program products implement embodiments of the present invention for data access that include identifying a set of tables in a database to be accessed by an application, and identifying first and second application programming interface (API) calls having different, respective access properties for accessing records in the tables via an API. Respective numbers of the records in the tables are counted by a processor, and the set of tables are partitioned into first and second subsets responsively to the respective numbers. The records in the first subset of the tables are accessed by having the application convey the first API call to the API, and the records in the second subset of the tables are accessed by having the application convey the second API call to the API.

Abstract: A computer-implemented method is disclosed for predicting a future usage of a cloud-based computing resource based on a previous usage of the resource by users, and predicting an anomaly event at the resource. The method also includes identifying a top contributing user responsible for the anomaly event, throttling an access of the top contributing user, evaluating a speed of data requests received from the top contributing user, and maintaining a utilization level of the resource within a predetermined target range. The method further includes dynamically controlling the speed of data requests based on the evaluation of the speed of data requests and a controlling speed of data request recommended by a first artificial intelligence model. The recommendations of the first artificial intelligence model may be validated by a human reasoning based model configured to monitor and mitigate a risk associated with a counter-intuitive recommendation of the first artificial intelligence model.

Abstract: Techniques are disclosed for managing metadata of a distributed database system in a hybrid manner. A computer system may receive, from a computing device, a request to access a set of data stored in nodes of a distributed storage system that is a caching layer of the system. The system retrieves metadata for a set of data specified in the request, including accessing a reversemap storing a reverse-ordered copy of original metadata stored in a metadata store of the system, where the reversemap is stored on durable storage of the system. Based on retrieving the metadata from the reversemap, the system accesses nodes of the distributed storage system, where the reversemap specifies the nodes of the distributed storage system that store the set of data. The system transmits, to the computing device, information indicating a result of accessing data stored in nodes of the distributed storage system.

Abstract: Techniques are disclosed for managing contention and improving transaction efficiency in distributed database systems through snapshot advancement. A system receives a query defining a first database transaction associated with a first snapshot of a database, where the transaction includes reading and updating a database value. The system evaluates the query against predefined criteria, such as accessing a single database row and deterministically locating the database value, to determine eligibility for snapshot advancement. Upon identifying contention over the database value between the first transaction and a second transaction, the system dynamically advances the transaction to a second snapshot of the database, avoiding a restart. This process involves granting snapshot advancement permission, waiting for a lock held by the second transaction to be released, and using the subsequent snapshot to continue execution.

Abstract: A multi-tenant system receives a request to build a software artifact for deploying on one or more target datacenter entities of the datacenter configured on the could platform. A software artifact comprises executable instructions associated with a service configured for execution on one or more cloud platforms. The system generates a release configuration that includes reusable release components and is used as a template by the release orchestration system to orchestrate release of the software artifact. The release configuration may specify one or more stagger groups, or a group of target datacenter entities that can be rolled-out in parallel.

Abstract: Using container information to select containers for executing models is described. A system receives a request from an application and identifies a version of a machine-learning model associated with the request. The system identifies a set of each serving container corresponding to the machine-learning model from a cluster of available serving containers associated with the version of the machine-learning model. The system selects a serving container from the set of each serving container corresponding to the machine-learning model. If the machine-learning model is not loaded in the serving container, the system loads the machine-learning model in the serving container. If the machine-learning model is loaded in the serving container, the system executes, in the serving container, the machine-learning model on behalf of the request. The system responds to the request based on executing the machine-learning model on behalf of the request.

Abstract: Techniques are disclosed relating to determining whether to utilize an alternative query plan for a subsequent execution of a database query. A computer system accesses a trace record detailing an execution of the database query in accordance with a first query plan selected by a query optimizer. The computer system may determine, based on the trace record, whether the execution of the database query satisfies a set of performance criteria. In response to determining that the execution of the database query does not satisfy the set of performance criteria, the computer system may determine whether one or more alternative query plans for the query include a second query plan that is predicted to be more performant than the first query plan. The computer system may provide an instruction to utilize the second query plan in place of the first query plan for the subsequent execution of the database query.

Abstract: Methods, systems, apparatuses, devices, and computer program products are described. A communication process flow management service may display a set of actions associated with a communication process flow configured to manage communications for a tenant. To update a segment of entities on which the communication process flow is to be executed, a user may navigate to a separate browser tab from that of the communication process flow management service, and may update the segment using an entity segmentation service. The communication process flow management service may receive an indication from an entity segmentation service that a segment of entities is updated and display a notification or a toast popup indicating the segment update to a user interface.

Abstract: A computing system may process entity name variations. The entity name variations may be transformed into vector representations. The vector representations may be stored in a database. The entity name variations may be clustered into groups of similar entity name variations based on a similarity metric between the vector representations. A canonical identifier for each group of similar entity name variations may be stored in the database. Each canonical identifier identifying a distinct entity.

Abstract: A method and apparatus for autonomous configuration-based release orchestration. A first engine obtains stagger configuration data that includes an indication of container clusters in each stagger and a stagger order, selects a current stagger based on the order, and attempts to deploy the feature release to the current stagger by causing an app config update to be sent to a second engine within each container cluster of the current stagger, and receives an indication of success or failure of the attempted deployment of the feature release to the current stagger. Responsive to the indication of success or failure, the first engine performs one of a plurality of actions that include attempting to deploy the feature release to a next one of the staggers according to the order responsive to the indication indicating success, and causing a roll back of the current stagger responsive to the indication indicating failure.

Abstract: Techniques are disclosed relating to determining query plans for execution by database systems. In various embodiments, a query optimizer determines a first query plan to implement a query requesting data from a database system. The determining includes selecting one of a plurality of query plans evaluated based on a cost analysis and caching plan fragments of the unselected query plans. The database system can then determine a second query plan for the query by replacing a plan fragment in the first query plan with one of the cached plan fragments of the unselected query plans.

Abstract: Techniques for providing a collaborative document in association with a virtual meeting on a communication platform are discussed herein. Users may communicate within the communication platform via text-based messages, which may include links, documents, videos, etc. Users may request a virtual meeting with audio and/or video capabilities. Responsive to a request, users within the communication channel may be invited to the virtual meeting and the collaborative document. The collaborative document can enable messaging and sharing functionalities attributable to a virtual space. The collaborative document may be populated with communications previously exchanged within the communication channel, such that links, documents, videos, content, etc. that occurred or were shared within the communication channel may be indicated within the collaborative document to provide context to the virtual meeting.

Abstract: Techniques and mechanisms described herein relate to a database system. A digital asset referenced in one or more database tables within an on-demand database system accessible to a plurality of entities may be identified. The digital asset may be owned by a first one of the plurality of entities. A smart contract that includes a token identifying the digital asset may be generated. The smart contract may be recorded within a distributed trust ledger. The token may be owned by a first account within the distributed trust ledger. The on-demand database system may include a correspondence linking the first entity with the first account. The on-demand database system may be updated to include a reference linking the digital asset with the token.

Abstract: Techniques are disclosed relating to database configuration settings overrides. In some embodiments, a database system stores a set of default configuration settings that control operation of the database system. The database system receives a query requesting data from the database system, and metadata about the query. The database system determines, based on the query and the metadata, that a configuration settings override has been specified for the query, where the configuration settings override indicates that one or more of the default configuration settings are to be replaced with one or more configuration settings specific to the query. In response to the determining that a configuration settings override has been specified, the database system executes the query using the one or more specific configuration settings.

Abstract: Apparatus and method for dynamic and persistent data sharing between cloud services. To address limitations in existing systems, resource provisioning and application deployment pipelines are configured with calls to share data using a configuration service. In some implementations, the configuration service includes an API, which is called by one or more stages of the resource provisioning pipeline to write configuration data in key-value pairs (e.g., data related to resource provisioning). One or more stages of the application deployment pipeline are configured with calls to access the configuration data. The stages then use the configuration data for application deployment. The configuration service manages the key-value store using versioning, repeatability, and immutability, to ensures that a configuration generated by any pipeline execution can be automatically re-used by another pipeline, at execution time.

Abstract: A computer system monitors for a change in a first data output generated by execution of a predefined dataflow. In accordance with a determination that the first data output has changed and the first data output meets triggering criteria, the computer system triggers execution of a predefined second dataflow distinct from the first dataflow. The execution of the second dataflow is dependent on the change in the first data output generated by the first dataflow. In accordance with a determination that the first data output has changed and the first data output does not meet triggering criteria, the computer system forgoes triggering execution of the second data flow. In accordance with a determination that the first data output has not changed, the computer system forgoes triggering execution of the second dataflow.

Abstract: Techniques described herein support component maintenance that accounts for a similarity between characteristics of different components. To identify components with similar characteristics, one or more techniques described herein support generation of a data structure (e.g., a tree) that represents a component, where each characteristic may be represented in a different leaf node. The system may generate a similarity score, or value (e.g., as a percentage), between multiple components based on comparing individual nodes of a tree representing each component, respectively. If the similarity score satisfies a threshold, then the system may display, at the UI, a message indicating to a user to preferentially implement one component over another.

Abstract: A conversational artificial intelligence (AI) point-of-sale (POS) system uses generative AI to help a store associate complete tasks in a retail environment. This POS system combines four technologies: (1) speech recognition; (2) natural language processing; (3) acting in response to verbal command in the context of a retail store and a POS device; and (4) providing bi-directional and fully conversational responses. The AI, which runs in an app on a smartphone, tablet, or other mobile device, uses a machine learning (ML) model, such as a small language model (SLM) or large language model (LLM), that is trained against support tickets and other product documentation. The AI understands dependencies (e.g., the tills should be reconciled before closing the store) and predict appropriate responses to conversational inputs.

Abstract: Examples include a system and computer-implemented method to receive a notification from an application programming interface (API) of creation of a just in time (JIT) grant, the JIT grant defining a request for a user to be authorized to access a cluster according to a JIT policy; determine if access to the cluster by the user is authorized according to the JIT policy; grant access to the user to the cluster when access is authorized according to the JIT policy; and send a notification to the API that access by the user to the cluster is granted.