Abstract: An online system allows user interactions using natural language expressions. The online system uses a machine learning based model to infer an intent represented by a user expression. The machine learning based model takes as input a user expression and an example expression to compute a score indicating whether the user expression matches the example expression. Based on the scores, the intent inference module determines a most applicable intent for the expression. The online system determines a confidence threshold such that user expressions indicating a high confidence are assigned the most applicable intent and user expressions indicating a low confidence are assigned an out-of-scope intent. The online system encodes the example expressions using the machine learning based model. The online system may compare an encoded user expression with encoded example expressions to identify a subset of example expressions used to determine the most applicable intent.

Abstract: Systems and methods for processing email messages are described. A method may include obtaining, from a database associated with the database system, data identifying a plurality of email messages for a plurality of email senders, the email messages associated with one or more sales cadences and an email service; enqueuing, by the server computing system, data identifying one or more email messages of the plurality of email messages into a queue provided that no data identifying two email messages associated with a first email sender are in the queue concurrently; and dequeuing, by the server computing system, the data identifying the one or more email messages from the queue, each dequeued data identifying an email message to be processed by the email service, wherein said enqueuing is performed provided that no dequeued data identifying two email messages associated with a second email sender are concurrently waiting to be processed by the email service.

Abstract: A group-based communication interface with subsidiary file communications is described. A request to share a file in a feed of a group-based communication platform can be received, wherein the feed comprises a channel, a thread, or a direct message associated with a group of the group-based communication platform. A representation of the file can be rendered in a user interface presented via client device(s) of user(s) associated with the group, wherein the representation of the file is associated with an engageable link for accessing the file, and wherein a thread summary indicator, representing a thread of message(s) associated with the file, is presented proximate the representation of the file. Based at least in part on receiving an indication of an interaction with the thread summary indicator, the message(s) associated with the thread can be rendered in the user interface proximate the representation of the file.

Abstract: A request for performing a data storing operation directed to a database table that comprises a plurality of table columns is received. Columnar compression metadata is accessed to identify one or more table columns in the database table, each of the one or more table columns being designated to store compressed columnar values. The columnar compression metadata is used to apply one or more columnar compression methods to generate, from one or more uncompressed columnar values received with the request for the data storing operation, one or more compressed columnar values to be persisted in the one or more table columns in the database table. A database statement is executed to persist the one or more compressed columnar values in the one or more table columns in the database table.

Abstract: Provided is a group-based communication interface configured to allow members of the interface to communicate within group-based communication channels. The group-based communication interface is configured to relegate selected group-based messaging communications to a separate display pane, where additional messaging communications can be received, thereby maintaining a focused selected group-based communication channel. The selected group-based communication channel is updated to indicate that the selected group-based messaging communications was relegated to a separate display pane, informing the channel members while also allowing the channel members to access the separate display pane if desired. Accordingly, the selected group-based communication interface provides for efficient and focused group-based communications.

Abstract: Methods, systems, and devices for configuring components in a display template based on a user interface type are described. A template customization application may support configuration of a display template corresponding to multiple user interface types. A user may customize the display template by adding a component to the display template (e.g., where the display template is displayed by the template customization application according to a specific user interface type view). The template customization application may determine whether the added component supports rendering for the current user interface type and may display the component in the template based on the determination (e.g., either as a functional component or as a placeholder). Each user interface type supported by the template may be defined in a single underlying metadata file, so adding the component into one view will automatically add the component for all of the supported the user interface types.

Abstract: Disclosed are methods, apparatus, systems, and computer readable storage media for interacting with a content object from an on-demand database service. The content object can be stored in an external content management data source. Access can be established with the external content object using information data identifying the content object, where the information data identifying the content object can be stored in a persistent object in a database of the on-demand database service. An indication of an event requesting interaction with the content object can be received. The persistent object can be updated to reflect the interaction with the content object.

Abstract: Systems, device and techniques are disclosed for training data generation for visual search model training. A catalog including catalog entries which may include images of an item and data about the item may be received. Labels may be applied to the images of the items based on the data about the items. The images of the items may be sorted into clusters using cluster analysis on the labels. Each cluster may include labels as categories of the cluster. Additional images may be received based on searching for the categories. Generative adversarial network (GAN) training data sets may be generated from the images of the items, the additional images, and the categories. GANs may be trained with the GAN training data sets. The GANs may generate images including images of generated items, which may be replaced with images of items from the catalog entries to create feature model training images.

Abstract: A system performs conversations with users using chatbots customized for performing a set of tasks. The system may be a multi-tenant system that allows customization of the chatbots for each tenant. The system processes sentences that may include negation or coreferences. The system determines a confidence score for an input sentence using an intent detection model, for example, a neural network. The system modifies the sentence to generate a modified sentence, for example, by removing a negation or by replacing a pronoun with an entity. The system generates a confidence score for the modified sentence using the intent detection model. The system determines the intent of the sentence based on the confidence scores of the sentence and the modified sentence. The system performs tasks based on the determined intent and performs conversations with users based on the tasks.

Abstract: Disclosed herein are system, method, and computer program product embodiments for determining the minimum set of user interface (UI) elements for a task. One embodiment operates by transmitting UI elements associated with an application to a first computing device operated by a first user and a second computing device operated by a second user. Further, the embodiment includes obtaining data structures corresponding to a first task and a second task and then grouping, using a machine learning model, the data structures into a first cluster corresponding to the first task and a second cluster corresponding to the second task. Further still, the embodiment also includes determining a minimum set of UI elements required for the first task based on the first cluster.

Abstract: Methods, systems, and devices for data validation are described. A user may store a set of data records on a source database and backup the set of data records at a target database through a data migration. A migration and validation server may initiate the data migration. After the data migration is complete, the migration and validation server may perform a validation process that includes comparing a calculated hash value from the source database and the target database that is based on unique identifiers and timestamps for each data record in the set of data records migrated from the source database to the target database. The migration and validation server may determine if the data migration was successful (e.g., the data was transferred correctly) if the hash value calculated for the data records at the target database equals the hash value calculated for the data records at the source database.

Abstract: A computing system uses neural networks to translate natural language queries to database queries. The computing system uses a plurality of machine learning based models, each machine learning model for generating a portion of the database query. The machine learning models use an input representation generated based on terms of the input natural language query, a set of columns of the database schema, and the vocabulary of a database query language, for example, structured query language SQL. The plurality of machine learning based models may include an aggregation classifier model for determining an aggregation operator in the database query, a result column predictor model for determining the result columns of the database query, and a condition clause predictor model for determining the condition clause of the database query. The condition clause predictor is based on reinforcement learning.

Abstract: Methods and systems are provided to determine suitable hyperparameters for a machine learning model and/or feature engineering process. A suitable machine learning model and associated hyperparameters are determined by analyzing a dataset. Suitable hyperparameter values for compatible machine learning models having one or more hyperparameters in common and a compatible dataset schema are identified. Hyperparameters may be ranked according to each of their respective influences on a model performance metrics, and hyperparameter values identified as having greater influence may be more aggressively searched.

Abstract: Methods, systems, and devices supporting data storage are described. A database server may store information in a group of files. As more information is stored at the database in the files, the sizes of the files may be increased. Techniques are described for uniformly growing files in the group of files to maintain a similar size for the files as the amount of storage increases. These techniques may prevent one file from becoming disproportionately larger than another file of the file group, supporting efficient read and write operations at the database. The file growth may be based on the file sizes as well as an amount of storage to be added for the group of files. For example, an application managing the file growth may sort the files by size, track uniform growth running totals, and determine file growth commands based on the uniform growth running totals.

Abstract: Methods, apparatuses, and computer program products for synchronization of live streams from web-based clients are disclosed. The method may include transmitting, to a plurality of clients broadcasting a plurality of media streams, an instruction to reproduce a synchronization signal. The method may include receiving, via the plurality of media streams, a plurality of recordings of the synchronization signal. The method may include determining one or more delays associated with the plurality of media streams based at least in part on the plurality of recordings of the synchronization signal. The method may include adjusting a timing of one or more of the plurality of media streams to synchronize the plurality of media streams, the adjusting based at least in part on the received plurality of recordings and the determined one or more delays. The method may include re-encoding the plurality of media streams into a combined stream.

Abstract: Methods and systems for securing customer data in a multi-tenant database environment are described. A key identifier received from a security server may be stored by an application server. The key identifier may be associated with a private key that is accessible by the security server and not accessible by the application server. A request to derive a symmetric key may be transmitted from the application server to the security server, the request including a public key generated by the application server, a salt value, and the key identifier. The symmetric key may then be derived based on the transmitted public key and the private key using a key derivation function. The application server may then receive and store the symmetric key in an in-memory cache, and be used to securely encrypt data received by the application server from client devices.

Abstract: Systems, apparatus, methods and computer program products for deleting data of an object within a multi-tenant database via an efficient delete operation are described herein. The new delete operation (or method) may be configured to identify one or more records to be deleted within a data structure such as an object. In order to ensure that the operation is efficient, the mechanism may determine characteristics of a data store and determine whether the information provided within the data structure provides an efficient identification of the data to be deleted. Upon a successful validation, an initial delete request may be translated to an appropriate delete operation for the underlying database.

Abstract: Methods, systems, and devices supporting virtual cryptographic key ceremonies are described. A server may receive a plurality of public keys and a plurality of digital signatures comprising data encrypted using a plurality of private keys, where each private key of the plurality of private keys corresponds to a respective public key of the plurality of public keys. The server may generate a quorum token based on the plurality of signatures and the plurality of public keys, where generating the quorum token is based on the plurality of signatures representing at least a threshold number of pools. The server may receive a plurality of encrypted shares associated with respective pools of a plurality of pools, generate a master wrapping key based on generating the quorum token and receiving the plurality of encrypted shares, unwrap a root key using the master wrapping key, and generate a certificate based on the root key.