Abstract: The technology described herein relates to systems, methods, and computer storage media, among other things, for generating recommendations corresponding to whether requests (e.g., transmitted by computing devices) are malicious network traffic. For example, the recommendations can be generated using various historical and current network traffic trends (e.g., associated with a particular application programming interface). As another example, the recommendations can be generated based on particular models. Based on the generated recommendation, dynamic rate limiting rules can be applied for determining whether a request is malicious network traffic. Based on determining the request is malicious network traffic, the request (and additional requests associated with that particular request) can be blocked.

Abstract: Query dependent threshold generation for search result filtering is described. In one or more implementations, a user query entered via a search platform is received, and in response, items are retrieved from a storage device based on the user query. Using a first machine learning model, relevance scores are generated for the items, and the relevance scores represent degrees of relevance of respective items with respect to the user query. Using a second machine learning model, a relevance threshold is generated for the user query based on one or more features of the user query. The items are filtered based on the relevance scores and the relevance threshold, and the filtered items are communicated over a network for display in a user interface of the search platform.

Abstract: Methods, systems and arrangements are described herein for enabling viewers to engage with live media broadcasts in ways that provide rich interactive experiences. Methods, systems and arrangements are also described for enabling transactions to be integrated with live broadcasts, and for enabling viewers to experience the feeling of being part of a group. Methods, systems and arrangements are also described for integrating viewer interactions into live media broadcasts themselves.

Abstract: Technology for the improved processing of search queries is provided. In one embodiment, methods may return semantically relevant search results for a search query. During a pre-computing offline processing, an inventory semantic index may be generated and may include inventory binary hashing signatures that are associated with inventory listings, such as goods or services for sell, and the index may be partitioned by categories and shards. When a search query is received, relevant categories are determined using a relevant category recognition service, and a search query binary hashing signature maybe generated for the search query. The relevant categories are searched to determine hamming distances between the inventory binary hashing signatures and the search query binary hashing signature, where the hamming distance indicates semantic relevance.

Abstract: Bitmask encoding-based personalized style generation is described. A personalization system presents different instances of digital content in a user interface and prompts a user to indicate interest or disinterest in individual digital content items. Based on the input indicating interest or disinterest, a bitmask encoding is generated and returned to the personalization system. The personalization system queries a database to identify respective instances of digital content that correspond to each binary value included in the bitmask encoding, which is then used to generate a prompt that causes a machine learning system to generate a personalized style for the user from which the favorable or unfavorable indications were received. The machine learning system generates a personalized style for an individual and identifies different instances of digital content of interest to the individual, and the personalized style is used to generate a user interface for the individual.

Abstract: Generating an access point certificate that defines relationships between an access point and at least one domain is described. Relationship data describing how data is to be routed between an access point and domains is received by a certificate management system. The certificate management system generates a certificate is generated that individually lists each domain associated with the access point and includes information describing data routing for the domain via the access point. The certificate excludes wildcard entries that represent multiple domains via a single entry. The certificate is used to control data communication traffic via the access point and is updated responsive to changes in in domain relationship data for the access point.

Abstract: AI-driven techniques are provided for generating multi-modal vector representations that incorporate enriched, model-generated category embeddings to enable conceptual search of item listings. For example, the enhanced conceptualized search can include generating a search vector, using one or more generative artificial intelligence (AI) models, based on a received search query. Some embodiments of generating the search vector may include applying one or more generative AI models to one or more sets of prior user interactions. Based on the search vector, a particular item listing vector can be determined. The item listing vector can be generated by applying one or more generative AI models to each of an image of the item and a textual description of an item listing, and the item listing vector can be generated based on enhanced categories determined (e.g., using one or more generative AI models) for the item listing.

Abstract: A housing includes an attachment mechanism, a light source including at least one light distributed around a first surface, and a spacer extending from the first surface by a first distance to maintain a fixed focal distance between a camera of a mobile device and an object to be imaged, the spacer forming at least a partially enclosed light shield that blocks external light.

Abstract: Music generation for item listings for an online marketplace is described. A user (e.g., seller) may list an item for sale on an online marketplace. The item listing may include information about the item, including a title, a description, and one or more images, among other information. The information may be input into a large language model (LLM), which may be trained on a set of music data and a set of inventory categories of the online marketplace. The LLM may execute to generate a music sample that corresponds to the item. The music sample may be applied to the item listing standalone, or as background music for a video corresponding to the item and included in the item listing.

Abstract: Automated image generation and optimization is described. In one or more implementations, a sample of listing titles is extracted from listings maintained by the online marketplace within a particular category. The extracted listing titles are simplified using at least one large language model (LLM) to remove extraneous information for image generation. An image prompt is generated using the LLM based on the simplified listing titles and the particular category. One or more images are generated using a large vision model based on the image prompt. The generated images are scored using a vision language model based on predefined criteria. The image prompt is iteratively refined based on the scoring, and images are regenerated until at least one image meets a threshold score for the predefined criteria.

Abstract: Some aspects relate to technologies for performing attribute-aware vector searches using adjusted similarity thresholds. In accordance with some aspects, a search query is received. Items are mapped onto an extended embedding space based on encoded attributes for the plurality of items. The search query is also mapped onto the extended embedding space based on encoded attributes for the search query. An adjusted similarity threshold is generated based on a number of attributes to be matched for the search query. Using the adjusted similarity threshold, items that are similar to the search query in the extended embedding space are retrieved. Search results based on the one or more retrieved items can be provided as a response to the search query.

Abstract: Some aspects relate to technologies for performing item retrieval on a listing platform using clusters of interchangeable parts formed using fitment data. In accordance with some aspects, item data is accessed for each of a plurality of part item listings on a listing platform, where the item data for each part item listing includes fitment data. An item embedding is generated for each part item listing using the item data. The item embeddings are clustered to generate a plurality of clusters, wherein each cluster includes one or more item embeddings. Cluster data is stored for the plurality of clusters. The cluster data for each cluster associates a cluster identifier and an item listing identifier for each item embedding in the cluster. The cluster data can be leveraged to perform item retrieval for the listing platform.

Abstract: In various example embodiments, a system and method for data mesh-based environmental augmentation are presented. Attribute data associated with a user may be received from a plurality of attribute sources. A portion of the attribute data may include real-time data. A portion of the real-time data indicative of an identity of the user may be identified. The identity of the user may be authenticated with respect to the real-time data by analyzing the identified portion of the real-time data. Based on the authentication of the identity of the user, a user activity being performed by the user may be identified based on the real-time data, and the user activity may be augmented according to a user setting.

Abstract: A visual search recommendation engine utilizes an image quality indication model for a visual search recommendation. Specifically, the visual search recommendation engine receives a search image as a search query at a search engine. The visual search recommendation engine provides the search image as an input into the image quality indication model, which is trained to output an image quality indication based on image aspects of the search image. A plurality of images are identified from an image corpus. The visual search recommendation engine determines an image similarity based on a comparison between the plurality of images from the image corpus and the search image. The image quality indication and the image similarity indicate a search query performance for the search image. A first image exceeding the search query performance is identified. The first image is provided for display at the search engine.

Abstract: Methods, systems, and computer programs are presented for replacing dialogue in a video segment. One method includes operations for analyzing content of a video to extract dialogue and meaning in the video, and for determining fragments of the video to be modified based on the extracted dialogue and meaning. The modification is based on factors comprising regional differences, cultural sensitivity, and inappropriate content. For each fragment to be modified, the following operations are performed: generate replacement speech based on the regional differences, cultural sensitivity, and inappropriate content; and generate audio for the replacement speech. The generation of the audio comprises synchronizing the audio with the video to align audio with lip movements while maintaining an emotional tone and a voice profile of each speaker in the video. Further, the method includes an operation for causing presentation on a computer display of the video with the modified one or more fragments.

Abstract: A large language model (LLM) based personalized chat quiz is described. In one or more implementations, information associated with an online marketplace is clustered into a plurality of topics. A plurality of prompts is generated, each prompt causing the LLM to generate a trivia quiz about a respective topic of the plurality of topics. Each prompt is generated by embedding a natural language description of the respective topic into a predefined prompt template. The prompts are input to the LLM, and generated trivia quizzes are received from the LLM. An indication of a customer service wait time is received. A generated trivia quiz about a personalized topic for the user based on tracked information about the user is selected from the generated trivia quizzes. The generated trivia quiz is embedded into the chat interface for presentation during the customer service wait time.

Abstract: In an example embodiment, a transaction log index is used to index log records. The transaction log index comprises an in-memory transaction log index and an on-disk transaction log index. The log records are temporarily indexed in the in-memory transaction log index until the storage server can ensure that the corresponding log records are stored without any missing log records in a sequence, as well as possibly based on other correctness and performance considerations.

Abstract: A real-time augmented reality item guide is described. A service platform may receive a digital image depicting at least one item. A plurality of listings for the at least one item on the service platform is identified via at least one machine learning model of the service platform based on the digital image, each of the plurality of listings associated with a respective point in time. An aspect is fetched from each of the plurality of listings for the at least one item on the service platform. Time-dependent augmented reality digital content is generated for the at least one item based on the aspect. The time-dependent augmented reality digital content is rendered as an overlay on the digital image.

Abstract: Automated determination of web page rendering performance techniques and systems are described. In one example, a web page including multiple user-interactive elements that can be selected or navigated via user input is received. A frame rendering system is employed to automatically identify the user-interactive elements and generate scripts to simulate user input at the user-interactive elements. The generated scripts are then used to simulate user input at the user-interactive elements. The scripts may be generated to simulate user interaction via a variety of different device types, and under a variety of different network conditions. During the simulation, a rendering performance score for the web page is generated by monitoring rendering performance relative to a threshold rendering rate.

Abstract: A method for constructing a point-in-time global consistent graph is described. The method includes accessing a stream of data from a transactional graph database, receiving, at a scalable persistent computational platform that is separate from the transactional graph database, a graph analytics query that indicates a time attribute value, in response to receiving the graph analytics query, constructing, at the scalable persistent computational platform, a point-in-time graph snapshot based on the stream of data and the time attribute value, and processing, at the scalable persistent computational platform, the graph analytics query with the point-in-time graph snapshot.

Abstract: Listwise autocomplete ranking with position debias is described. A computing system can receive a user input indicative of a portion of a search query. The computing system can receive a set of candidate queries for an autocomplete list based on the portion of the search query. The computing system can generate a respective ranking score for each candidate query of the set of candidate queries by calculating, for each candidate query of the set of candidate queries, a respective click probability prediction using a position debiased listwise click-through rate (CTR) prediction model, and calculating, for each candidate query of the set of candidate queries, a respective reward per search. The computing system can order the set of candidate queries according to the respective ranking scores. The computing system can output at least one of the candidate queries based on the ordered set of candidate queries.

Abstract: In accordance with the described techniques, a configuration update is received changing configuration parameters of a data stream processing application from a source configuration to a target configuration. In addition, the configuration update specifies incremental update parameters indicating how different partitions of the data stream processing application are to be incrementally updated. The computing device is configured to incrementally update the data stream processing application. During a time interval indicated by the incremental update parameters, for instance, the target configuration is applied to a first partition, and the source configuration is applied to a second partition. Thus, during the time interval, the first partition processes events of a data stream in accordance with the target configuration, and the second partition processes events of the data stream in accordance with the source configuration.

Abstract: Techniques are described, as implemented by computing devices, to control access to transactions through use of security deposits based on tokenized reputation scores. This is performed by leveraging a blockchain such that tokenized reputation scores are generated based on amounts of cryptographic reputation tokens associated with blockchain account addresses associated with service provider accounts. Access to transactional functionality is controlled for a service provider account by using an amount of cryptographic reputation tokens as a security deposit, where the security deposit is calculated based on a tokenized reputation score of the service provider account. Further, return of a security deposit is controlled for a service provider account based on a determination of a successful or unsuccessful completion of a transaction.

Abstract: The present disclosure is directed to providing supplemental content to one or more client devices requesting multimedia content. The supplemental content may be determined or selected based on the availability of one or more device channels, such as a display, speakers, or other component of the one or more client devices capable of providing an output. Furthermore, the supplemental content may be determined or selected based on the portions of the requested multimedia content that are the most prominent or significant, such as any audio content, any video content, and/or any textual content. The supplemental content may be provided to the one or more client devices such that it is displayed before, during, or after the display of the requested multimedia content.

Abstract: Some aspects relate to technologies for providing keyboard emulators on product webpages of mechanical keyboards to emulate the tactile and/our auditory features of the mechanical keyboards. To configure a keyboard emulator for a given mechanical keyboard, a user interface is provided for presentation on a first user device. An input is received via the user interface that adjusts an element of a haptic pattern and/or a sound pattern for the keyboard emulator. Based on the input, a configured haptic pattern and/or configured sound pattern is provided for the virtual keys of the keyboard emulator. Once configured, the keyboard emulator is provided on a product webpage for the mechanical keyboard for presentation on a second user device. User interaction with virtual keys of the keyboard emulator causes the second user device to provide a haptic feedback and/or sound using the configured haptic pattern and/or configured sound pattern.

Abstract: Described are computing systems and methods configured to detect a small, but meaningful, anomaly within one or more metrics associated with a platform. The system displays visuals of the metrics so that a user monitoring the platform can effectively notice a problem associated with the anomaly and take appropriate action to remediate the problem. An operational visual includes a radar-based visual with a heatmap arranging metrics, and a node representing a state of the metrics. Moreover, the system uses an ensemble of unsupervised machine learning algorithms for multi-dimensional clustering of hundreds of thousands of monitored metrics. Via the visuals and the implementation of the machine learning algorithms, the described techniques provide an improved way of representing and simulating many metrics being monitored for a platform. Moreover, the techniques are configured to expose actionable and useful information associated with the platform in a manner that can be effectively interpreted.

Abstract: An immersive trading card experience is described. In one or more implementations, an image of a front of a trading card is obtained and segmented into two or more layers displayable at different depths. Listing information corresponding to the trading card is retrieved for use as the trading card rear and is adapted to include real-time trending information for additional listings of the trading card. In response to the trading card being returned in a search result, an enhanced item card is communicated to a computing device. The enhanced item card includes the trading card front with the layers displayable at different depths. A rotatable display of the enhanced item card includes the layers displayable at different depths as an enhanced image of the trading card front and includes the listing information as the trading card rear.

Abstract: Key phrase recommendation based on token correspondence is described. In one or more implementations, a plurality of user queries entered via a search feature of a listing platform is received. Key phrases are extracted from the plurality of user queries. The key phrases are associated with an item category of the listing platform, and each have a search count that exceeds a threshold. Further, an item title of an item listed via the listing platform in the item category is obtained, and the item title includes title tokens. The key phrases are ranked based, in part, on quantities of the title tokens included in the key phrases, and the ranked key phrases are output.

Abstract: A system for automatically generating auto-scaling configurations for graphics processing unit (GPU) models is described. A provides users a platform to perform a load and performance (LnP) test on a model of a GPU of a computing device. The LnP test may result in a set of metrics associated with the GPU model. For example, the metrics may include throughput metrics (e.g., transactions per second (TPS)) and GPU utilization metrics. In some examples, the metrics may be based on a service level agreement (SLA), which may include latency, error rate, and GPU utilization requirements. Based on a scaling threshold determined from the metrics and a utilization requirement, the system may output an auto-scaling configuration for the GPU model. The GPU may operate using the auto-scaling configuration, where the auto-scaling configuration may enable the GPU model to scale up or scale down, for example, based on changes in traffic.

Abstract: An artificial intelligence (AI)-based system for component compound identification is described. The system may receive images associated with a product listing for an online marketplace. The product listing may be for a medicinal substance (e.g., a prescription drug). The system may generate a prompt for a large language model (LLM) based on the images and prompt the LLM to identify component compounds (e.g., ingredients) of the medicinal substance. For example, the LLM may recognize text from the images that indicates the component compounds and their respective amounts (e.g., weights or percentages). The system may generate an output based on prompting the LLM, the output indicating the identified component compounds and a determination of whether the product listing is approved or denied.

Abstract: Controlling access to at least one domain associated with an access point using a wildcard-free certificate and an allow list is described. Relationship data describing how data is to be routed between an access point and domains is received. Based on the relationship data, a certificate is generated that individually lists each domain associated with the access point and includes information describing data routing for the domain via the access point. Data describing at least one exception for granting a request, when the request does not specify a domain included in the wildcard-free certificate, is received and used to generate an allow list. The certificate and the allow list are used to control data communication traffic via the access point.

Abstract: Technologies are shown for shipping route selection involving receiving sender and recipient shipping information for an item to be shipped and obtaining shipping route options for the item based on the sender and recipient shipping information. The route options are provided for display and selection of an option. A routing data block is created for the item at an address on a blockchain that stores shipping information for each stage of the selected route. A shipping tag is encoded with the blockchain address and attached to the item. The tag can be scanned to obtain the blockchain address and request information for a next stage of the shipping route from the block. The next shipping information from the block is received and utilized to ship the item to a next geolocation. The next shipping information can be determined based on current conditions, such as weather, pricing and availability.

Abstract: In various example embodiments, a system and method for dynamically generating user interface elements and associated values are presented. An item listing and profile data are accessed. A set of user interface elements are dynamically generated based on the item listing and the profile data, with each user interface element configured to perform an action on the item listing. A set of values are dynamically determined, with each value being associated with a user interface element of the set of user interface elements. The set of user interface elements are then caused to be presented within the item listing.

Abstract: In implementations of systems and procedures for item recommendation and visualization, a computing device receives an input digital image depicting an environment and identifies attributes of objects within the environment. The attributes of the objects are used to identify items from item catalog data that have similar attributes and are suitable for inclusion within the environment. The identified items are displayed by way of a recommendation. The attributes of the objects are further used for generation of synthesized images. The synthesized images depict objects from the input digital image virtually replaced by items from the item catalog data. The synthesized images thus support visualization of the items within the environment depicted by the input digital image.

Abstract: Detecting imbalances in data traffic for distributed data centers is described. A distributed data center system can obtain a set of data samples corresponding to respective data traffic at data centers, the respective data traffic based on data associated with an application. The distributed data center system can generate subsets of weight-transformed data samples associated with respective intervals of time. The distributed data center system can generate respective indices associated with the subsets of weight-transformed data samples. The indices can be representative of a data traffic balance at the data centers. The distributed data center system can transmit a message indicating an index failing to satisfy a threshold value, the index being associated with at least one subset of weight-transformed data samples.

Abstract: In the implementation of techniques for generating guidance digital content with artificial intelligence, a system receives digital content and metadata corresponding to item listings for an item category. The system filters the digital content based on predefined quality metrics and metadata to identify a subset of reference digital content, where each reference digital content meets or exceeds a predefined quality threshold. Via one or more artificial intelligence models, the system classifies the subset of reference digital content into distinct perspectives, each distinct perspective representing a different view of items included in the item category. Based on this classification, via the one or more artificial intelligence models, the system generates guidance digital content for each distinct perspective. The system broadcasts the guidance digital content for display via a user interface in association with the item category.

Abstract: Methods, apparatus, and systems to determine a niche market of items or services, the first phase of which identifies a gap between demand and supply for a set of items. Session logs may be evaluated to compare transactions involving a specific item to those of a larger group of items. The resultant information identifies areas of high demand, but with low availability. The niche market information may be provided as direct merchandising items for sellers. In one example, the method generates niche market item web pages in specific categories. Additional methods, apparatus, and systems are disclosed.

Abstract: Listwise autocomplete ranking with position debias is described. A computing system can receive a user input indicative of a portion of a search query. The computing system can receive a set of candidate queries for an autocomplete list based on the portion of the search query. The computing system can generate a respective ranking score for each candidate query of the set of candidate queries by calculating, for each candidate query of the set of candidate queries, a respective click probability prediction using a position debiased listwise click-through rate (CTR) prediction model, and calculating, for each candidate query of the set of candidate queries, a respective reward per search. The computing system can order the set of candidate queries according to the respective ranking scores. The computing system can output at least one of the candidate queries based on the ordered set of candidate queries.

Abstract: Some aspects relate to technologies for generating pivots using a generative model and grouping items into pivot groups using the pivots. Pivots are generated by obtaining item information for items, generating a prompt using the item information, and causing a generative model to use the prompt to generate text for the pivots, including a pivot name and pivot description for each pivot. A pivot embedding is also generated for each pivot. When items are to be returned to a user device (e.g., as recommended items or search result items), item embeddings for the items and the pivot embeddings for the pivots are used to assign each item to a particular pivot to generate pivot groups that each includes a pivot name, a pivot description, and assigned items. User interface information is provided to present at least a portion of the pivot groups on a user device.

Abstract: Example methods and systems are directed to a managed inventory. A database may store information regarding items owned by a user. The information regarding an item may include a quantity owned and one or more triggering events. Based on the occurrence of a triggering event, an order for the item may be placed without user intervention. Data to the database may be provided by one or more sensors. Triggering events may be defined in terms of sensor data. The triggering event may be defined by a user or through machine learning. The order may be placed using a predetermined modality or a dynamically-determined modality based on one or more criteria, such as price, shipping speed, and the urgency of the order.

Abstract: Rules based override of machine learning output is described. In one or more implementations, the described architecture receives a listing for an item having one or more attributes describing the item. A category of the item, as output by a machine learning model based on the listing, is also received. The machine learning model is trained using a dataset of listings labeled with item categories. A set of deterministic rules is executed on the one or more attributes describing the item and the output of the machine learning model, including identifying the category of the item and applying one or more of the deterministic rules associated with the category. Based on the executing, a different category of the item is output and overrides the category of the item as output by the machine learning model.

Abstract: Dynamic automatic generation of item listings is described. A computing device (e.g., or a user engagement system) receives input that indicates at least one item. The computing device generates a probability of user engagement with a listing of the at least one item. The computing device generates the probability based on a context associated with the at least one item. In some cases, the computing device displays a control selectable to automatically generate the listing of the at least one item based on the probability of the user engagement satisfying a threshold value. In some other cases, the computing device displays a control selectable to generate a configurable template for the listing of the at least one item based on the probability of the user engagement failing to satisfy a threshold value.

Abstract: A computer-implemented method and system provides audio and language translation between a speaker at a first computing device and a listener at a second computing device. The first computing device inputs speech in the speaker’s language pre-defined as corresponding to the speaker, translates the speech from the speaker’s language into audio data in the listener’s language predefined as corresponding to the listener. The first computing device superimposes the speaker’s pronunciation as modeled by a speaker pronunciation model onto the audio data in the listener’s language so that the pronounced audio data in the listener’s language will sound as if it is spoken by the speaker. The speaker pronunciation model is trained on the speaker’s voice speaking the speaker’s language and remains stored at the first computing device. The pronounced audio data is streamed to the second computing device while the speaker at the first computing device is speaking.

Abstract: A method for detecting an anomaly is provided. The method receives and organizes data as a function of a plurality of time periods and a number of updates for each time period of the plurality of time periods. The organized data approximates first and second unimodalities. The first and second unimodalities have respective first and second peaks. The method identifies a dip across the plurality of time periods and determines a density of the dip. The density is compared with a predetermined threshold in order to determine if the dip is an alternative modality. A unimodality analysis is then performed on the first and second unimodalities in response to the dip being an alternative modality. The method also determines whether the second unimodality is an anomaly based on the unimodality analysis and discards the updates associated with the second unimodality when the second unimodality is an anomaly.

Abstract: Some aspects of the present technology relate to technologies for context and knowledge sharing abstraction across entities using a relational hierarchical key mapping system. In accordance with some configurations, a first set of one or more data points for a first set of one or more keys is received at a first domain and stored in a schema-less data store. A second set of one or more data points for a second set of one or more keys is received at a second domain and stored in the schema-less data store. Based on at least one shared key, the first set of one or more data points and the second set of one or more data points are linked as a virtual entity such that a request to retrieve data points of the one or more data points is performed as a single seek and a single API call.

Abstract: Various methods and systems for providing indications of inconsistent attributes of item listings associated in item listing videos. An item listing video—of an item listing—is accessed. The item listing video is accessed via an item listing interface of an item listing system. Extracted item features—via a machine learning engine—of an item from the item listing video, are accessed. The extracted item features are extracted based on listing-interface item features associated with listing the item. The extracted item features of the item are compared to the listing-interface item features of the item. Based on comparing the listing-interface item features to the extracted, an inconsistent attribute—between an extracted item feature and a listing-interface item feature that is associated with listing the item—is identified. An indication of an inconsistent attribute is communicated to cause display of the indication of the inconsistent attribute at the item listing interface.

Abstract: Digital content authentication using a cryptographic possession factor is described. In one or more examples, a navigation request to access a webpage is received and verified, over a plurality of iterations over a session, that an originator of the navigation request maintains a cryptographic token as bound within a browser through use of a public key associated with the cryptographic token. A log describing the verifying is stored and then a determination is made that the webpage involves additional authentication usable to extend the session to access the webpage. A level of the authentication is selected for access to the webpage based on the cryptographic token maintained at the browser and the log. Access to the webpage is controlled based on the selected level of authentication.

Abstract: An artificial intelligence (AI)-based smart actioning system is described. A system receives an activity associated with a website, the activity having one or more attributes. Based on the attributes, the system generates a vector associated with the activity. The system executes a machine learning model that is training using a dataset including a vector space of a set of vectors associated with a set of historical activities associated with the website. This may include executing a vector space similarity search, such as a k-nearest neighbor (k-NN) search, between the vector and the set of vectors. The system generates a prompt for a large language model (LLM) based on one or more vectors identified from the vector space similarity search, and outputs a result received from the LLM that indicates a risk level of the activity.

Abstract: Technologies are disclosed herein for distributing information. The disclosed technologies determine an application element configured to receive information that is updated at a variable rate, the information pertaining to an object. Feature data is received that is associated with the object and data associated with use of the application element. The feature data includes a time horizon for the object and supplemental information associated with the object. Based on the feature data and the data associated with use of the application element, a first rate is predicted for sending the information about the object to the application element. The information is sent to the application element at the first rate.

Abstract: Techniques are described, as implemented by computing devices, to control access to transactions through use of cosigning based on tokenized reputation scores. This is performed by leveraging a blockchain such that tokenized reputation scores are generated based on amounts of cryptographic reputation tokens associated with blockchain account addresses associated with applicant and co-signer service provider accounts. Transactional functionality is made available to an applicant service provider account having an insufficient tokenized reputation score by using a co-signer service provider account having a sufficient tokenized reputation score to at least partially back an obligation of the applicant service provider for a transaction.