36 results about "Request - action" patented technology

This disclosure describes a framework for performing user-requested tasks automatically across an interactive interface using various types of machine learning models. Specifically, this disclosure outlines and describes a task execution system that utilizes a generative artificial intelligence (AI) action model and retrieval-augmented generation (RAG) to complete user-requested actions across an interactive interface. The task execution system solves many of the current limitations of LAMs by using a generative AI action model to determine a session plan, which includes a set of actions for accomplishing stages of the actionable task across the interactive interface, obtaining visual context information of each interactive interface segment, integrates RAG results to improve the accuracy of both the session plan and individual actions, and self-corrects when faced with unexpected obstacles.

A system for conserving computing and memory resources and enhancing security by performing application-level multifactor level authentication of a client requesting execution of an action by the system. A request to execute an action sent by the client and received by an action execution application of the system may include compiled client authentication information. The compiled client authentication information may comprise at least a client digital certificate and client login credentials that are unique to the action execution application. The action execution application can authenticate the client relative to the requested action by validating the received client digital certificate against a stored set of previously validated digital certificates, in addition to validating the received client login credentials against a stored set of previously validated login credentials. In response to authenticating the client, the action execution application may execute the requested action.

Techniques for masking data based on annotations are discussed herein. A system may receive a request to perform an action and leverage an LLM to assist in performing the requested action. When generating the input data to input to the LLM, the system can use a template to organize the input data. The template may include static data and / or slot(s) which can include a reference to data to input into such slots. The system may retrieve data to input to the slot based on the reference, retrieve annotations that define a classification of the data, and receive a policy that defines which types of data classifications are to be masked. Based on the data classification and the policy, the system can determine whether to mask the data. The system can generate the input data using the template, the data, and / or the mask(s) and input such data into the LLM.

Techniques for streaming generative machine learned model (or LLM) output to a virtual space are described herein. A system may receive a request to perform an action. The system may leverage LLMs to assist in performing aspects of the requested action. The system can generate input data to input into the LLM. When generating the input data, the system can identify sensitive data associated with the request. The system can modify the input data to mask and / or anonymize the sensitive data. The system can input the input data into an LLM trained to output a subset (e.g., less than all) of the response at a time. The system can add the output subset to a buffer and upon identifying the masked data in the buffer, the system can demask the sensitive data and output the sensitive data to the user profile in a streaming manner.

A wireless transmit / receive unit (WTRU) may receive a first message broadcast by a network. The first message may indicate one or more challenge values, an estimated required energy level for performing the requested action, and / or an indication that the WTRU should respond with an indication of an amount of energy available to the WTRU. The WTRU may determine whether to respond to the first message based on the amount of energy available to the WTRU. The WTRU may send a second message to the network based on the determination to respond to the first message. The second message may indicate a challenge response value, an amount of energy available to the WTRU, and / or a device-specific network challenge value.

The present disclosure relates to systems, non-transitory computer-readable media, and methods for processing access requests on a service-to-service basis using a third-party identification token. In particular, the disclosed systems can identify that a user is authenticated for a first computer service based on detecting a third-party identification token. Further, the disclosed systems can generate, by the first computer service, an access request comprising a requested action and the third-party identification token to a second computer service. Additionally, the disclosed systems can determine, by the second computer service, whether the access request is authorized based on determining that an authorization policy defined at the second computer service authorizes the requested action by the first computer service and that the third-party identification token is valid. Moreover, the disclosed systems can provide, to the first computer service, a response to the access request in response to determining whether the access request is authorized.

Systems and methods of selecting digital platforms for execution of voice-based commands are provided. The system receives an application that performs an action associated with a service via digital platforms. The system debugs the application to validate parameters of the action on at least two platforms of the digital platforms. The system receives data packets comprising an input audio signal detected by a sensor of a client device, and parses the input audio signal to identify the action and the service. The system selects a first platform from the digital platforms to perform the action. The system initiates, responsive to selection of the first platform, an interactive data exchange to populate parameters of an action data structure corresponding to the action. The system executes the action via the selected platform using the action data structure.

One wireless communication method includes: receiving a request action frame; and in response to the request action frame, generating and sending an acknowledgement (ACK) control frame that is configured to serve as a response action frame for the request action frame, wherein the response action frame is not solicited by the request action frame. Another wireless communication method includes: receiving a request action frame; and in response to the request action frame, generating and sending an acknowledgement (ACK) control frame and a time-constrained response action frame following the ACK control frame, wherein the time-constrained response action frame is solicited by the request action frame.

This relates generally to intelligent automated assistants and, more specifically, to provide intelligent device selections by the intelligent automated assistants for performing requested actions. An example method includes, at an electronic device receiving a user request from a user, identifying the user, a domain type of the user request, and one or more electronic devices available for handling the user request; retrieving one or more historical interactions involving at least one of the identified user, the domain type, and the one or more electronic devices, generating metadata based on the one or more historical interactions, location information of the one or more electronic devices, and context information associated with the one or more electronic devices; identifying a delivery device by interpreting the metadata using a preference model; and transmitting a response command to the delivery device for providing the result output.

A system and method for authentication of a requested action for account security are provided. The method may include receiving notification of the requested action associated with an account. The method may include transmitting an end-user prompt to a prompt endpoint associated with the account. The method may include receiving a response message including an end-user response which may indicate approval of the requested action from the prompt endpoint and which includes data points collected from the prompt endpoint. The method may include processing the collected data points to determine an updated authentication activity requirement, which may include a requirement to decline or further verify the requested action despite the end-user response message indicating approval of the action.

Various embodiments of the present technology generally relate to systems and methods for providing an authentication engine for monitoring interactions associated with network functions. In an example, an authentication engine may determine a request from a client device that includes a requested action associated with a network function within a network. The client device may have authentication credentials for performing the requested action. The authentication engine may determine a deviant intent based on the requested action, such as determining the requested action being outside a range of normal activity and / or associated with an active network function. Based on the deviant intent, the authentication engine may generate a subsequent authorization requirement that is required to be completed for the client device to perform the requested action.

Methods and systems are described for utilizing machine learning models to generate a response. The system receives an operator input requesting actions to prompt a desired user interaction. The system selects user-specific machine learning models trained on interactions associated with each user of a plurality of users. Based on the operator input, the system determines a model subset and selects a candidate operator action to test. The system inputs the candidate action into each model of the subset to obtain a predicted user interaction. The system then assesses whether the candidate action causes the models to predict the desired interaction. If the candidate action is effective, the system executes the candidate action in relation to the user subset.

A method for controlling a remote control device includes capturing with a biometric device biometric data associated with a user, determining with processor whether the user is authorized to interact with the smart device, in response to the biometric data, determining with the processor user data, in response to the user being authorized to interact with the smart device, receiving with a short-range transceiver an authentication request from a reader device associated with a remote control device, outputting with the short-range transceiver a session token in response to the authentication request and to the user data, thereafter determining with a physical sensor physical perturbations in response to physical actions of the user, determining with the processor a requested action for the remote control device, in response to the user data and the physical perturbations, and outputting with the short-range transceiver the requested action to the reader device.

Two party computer access control including a network interface for receiving, an initiate message including a first label, a first payload and a first internet protocol address and for receiving a control message including a second label, a second payload and a second internet protocol address, and a processor for executing the action in response to the first label corresponding to the action, the first internet protocol address corresponding to an expected internet protocol address of the first user system, the first payload corresponding to a first expected payload associated with the requested action, the second label corresponding to the action, the second internet protocol address corresponding to an expected internet protocol address of the second system, and the second payload corresponding to a first expected payload associated with the action.

An information handling system for an on-the-box artificial intelligence productivity tool may comprise a memory storing descriptions of capabilities associated with software applications in a decision tree with each capability and a generated capability intent value stored as a capability node grouped under one of several branches in parent-child relationships, and a hardware processor executing machine readable code instructions to generate a query input intent value from a user query input requesting action by one of the software applications, comparing the capability intent values of each of the capability nodes in the decision tree, as weighted by a lexical comparison between the user query input and each of the descriptions of capabilities to conduct a breadth-first search to identify a best match childless capability node having a highest lexical weighted cosine semantic similarity search score, and executing the associated best match capability from the best match childless capability node.

Systems and methods are provided for securely activating actions by work machines, independent of distance between the machines and users requesting the actions. An authentication network is established between a host server, authorized users, and work machines, wherein authorized user computing devices are assigned authentication keys. A user sends a first message to the host server comprising requested actions by an identified work machine and signed with an authentication key. The host verifies authenticity of the first message, optionally adds a host signature, and transmits the verified message to the work machine. The machine controller verifies authenticity of the first message and conditionally generates a second message comprising a confirmation challenge to the user. If the user responds satisfactorily to the challenge, the machine executes the requested actions, for example relating to startup / shutdown, access control, autonomous control of operations, adjustments to machine settings, and the like.

The electronic system is configured to process a natural language spoken (720) by a computer game player during game play using a customized large language model (740). Then, in response, a non-verbal output may be provided (750), where the non-verbal output may be visual, tactile, and / or auditory. As an example, non-verbal output may be responsive to a request from a user, may confirm that a natural language is being processed, and / or may indicate that a requested action has been registered for future execution.

A wireless communication method includes: receiving a request action frame; and in response to the request action frame, generating and sending an acknowledgement control frame that is configured to serve as a response action frame for the request action frame, wherein the response action frame is not solicited by the request action frame. A wireless communication device (102) includes a network interface circuit (118) and a control circuit (116). In response to a request action frame received by the network interface circuit (118), the control circuit (116) generates an acknowledgement control frame that is configured to serve as a response action frame for the request action frame, and instruct the network interface circuit (118) to send the acknowledgement control frame, wherein the response action frame is not solicited by the request action frame.

A system for conserving computing and memory resources and enhancing security by performing application-level multifactor level authentication of a client requesting execution of an action by the system. A request to execute an action sent by the client and received by an action execution application of the system may include compiled client authentication information. The compiled client authentication information may comprise at least a client digital certificate and client login credentials that are unique to the action execution application. The action execution application can authenticate the client relative to the requested action by validating the received client digital certificate against a stored set of previously validated digital certificates, in addition to validating the received client login credentials against a stored set of previously validated login credentials. In response to authenticating the client, the action execution application may execute the requested action.

An agent computer system uses a session-less login process to log in two users in a same application session. The system establishes an application session with an application server and a user session with the application server by authenticating a first user using an identity provider system (IdP). The system detects a request to perform an action by the first user within the application session that requires a second user's authentication. While maintaining the user session, the system requests a session-less authentication of the second user by transmitting an authentication request, which includes a flag indicating that the requested authentication is a session-less authentication. The system receives a confirmation of the authentication of the second user without establishing a second user session and performs the requested action in response to receiving the confirmation of the authentication of the second user.

An OTN network electric layer service dynamic route allocation method, apparatus and device, and a computer readable storage medium, the method comprising: training a main neural network according to constructed training set data to obtain target parameters of the trained main neural network, the training set data comprising multiple groups of training data, and the training set data comprising multiple groups of training data; the training data comprises a current global network state, a service request action, a return value and a next-moment global network state; updating parameters of a target neural network in an OTN network controller according to the obtained target parameters to generate a target model; outputting a target value of each path according to an obtained new service request based on the target model; and determining a target path according to the target value of each path, and allocating bandwidth resources to the target path, thereby solving the technical problem that the balance of service transmission delay, channel utilization rate and channel resources is difficult to optimize at the same time in the prior art.

A computer-implemented method for providing location-based appointment operations is disclosed. The method includes receiving input indicating an instruction to perform an action related to appointments. Responsive to the input indicating the instruction to perform the action related to appointments, image recognition techniques may be used to determine that a computing device is proximate to a physical location where locations are schedulable. Then, responsive to determining that the computing device is proximate to the physical location, information related to scheduling appointments at the physical location is received via a network. The indication based on the received information is presented at the computing device. Input requesting an action in relation to an appointment at the physical location may be received at the computing device. An indication to initiate the requested action may be sent via the network. Related computing devices and computer-readable media are also disclosed.

Methods and systems for routing execution requests to autonomous artificial intelligence agents using performance-based selection are disclosed herein. An agent routing system may receive a prompt for completing an action using artificial intelligence agents from a plurality of agents, where each agent may be associated with computer-executable operations configured for autonomous execution on software applications. The system may determine a set of agents enabled to complete the requested action and may access action completion datasets for these agents. Each dataset may include action parameters for completed actions, including status parameters indicating successful completion and timestamps of completion. The system may select an artificial intelligence agent based on a frequency composite parameter generated from successful completion status parameters and a time composite parameter indicating completion timing. The system may identify parameters required by the selected agent to complete the action and may transmit these parameters to the artificial intelligence agent for execution.

An information handling system for an on-the-box artificial intelligence productivity tool may comprise a memory storing descriptions of capabilities associated with software applications in a decision tree with each capability and a generated capability intent value stored as a capability node grouped under one of several branches in parent-child relationships, and a hardware processor executing machine readable code instructions to generate a query input intent value from a user query input requesting action by one of the software applications, comparing the capability intent values of each of the capability nodes in the decision tree, as weighted by a lexical comparison between the user query input and each of the descriptions of capabilities to conduct a breadth-first search to identify a best match childless capability node having a highest lexical weighted cosine semantic similarity search score, and executing the associated best match capability from the best match childless capability node.

An information handling system managed by an enterprise management system and operating an on-the-box artificial intelligence productivity tool may comprise a hardware processor executing machine readable code instructions to generate vectorized capability intent values from natural language descriptions of capabilities including a capability-with-reboot requiring reboot to a basic input / output system (BIOS) and a vectorized query input intent value for a user query input requesting action from information handling system, compare the query input intent value to the capability intent values to identify a best match responsive capability-with-reboot for the received user query input having a capability intent value that generates a highest semantic similarity search score and requiring authorization for execution of pre-boot machine readable code instructions for the best match capability-with-reboot received from the enterprise management system, store it in a pre-boot memory partition, reboot, and execute the best match responsive capability-with-reboot upon boot to BIOS.

Implementations relate to interactions, over a wireless communication modality, between an automated assistant of an automobile computer system, and separate client device(s). In some of those implementations, the client device(s) can be paired with the automobile computer system, and the client device(s) and the automobile computer system can include separate instances of an automated assistant, which can be associated with different user accounts. In some additional or alternative implementations, particular actions to be performed by a local application of a client device can be advanced via user interaction with the automobile automated assistant, despite a counterpart to the local application not being installed directly on automobile computer system. For example, despite an automobile computer system not having a third-party messaging application installed, the automobile automated assistant can access requested actions to be performed by the third-party messaging application in order to further a requested action. The automobile computer system can transmit, via the wireless communication modality, content to cause the third-party application to further the requested action.

A computer-implemented method for activity monitoring with respect to online services. The method can include accessing a machine learning multiclass classifier, the machine learning multiclass classifier representing HTTP network request features and associated actions with respect to interacting with websites, receiving an HTTP request, extracting a feature set from the HTTP request, determining a request action classification for the HTTP request, determining the request action classification comprising processing the feature set extracted from the HTTP request to the machine learning multiclass classifier to classify the HTTP request, and providing access to the HTTP request and request action classification via an application programming interface.

A method for controlling a remote control device includes capturing with a biometric device biometric data associated with a user, determining with processor whether the user is authorized to interact with the smart device, in response to the biometric data, determining with the processor user data, in response to the user being authorized to interact with the smart device, receiving with a short-range transceiver an authentication request from a reader device associated with a remote control device, outputting with the short-range transceiver a session token in response to the authentication request and to the user data, thereafter determining with a physical sensor physical perturbations in response to physical actions of the user, determining with the processor a requested action for the remote control device, in response to the user data and the physical perturbations, and outputting with the short-range transceiver the requested action to the reader device.

This invention provides an electronic device and a method for waking the electronic device from sleep mode. When the electronic device is in sleep mode, if it receives an intermittent low-power threshold detection enable signal, it enters a low-power threshold detection mode. When the electronic device is in sleep mode, if it receives a sampling request action signal or acceleration data exceeding a first threshold range, it enters a continuous operating mode. If it does not receive a sampling request action signal within a predetermined time period or the acceleration data does not exceed a second threshold range within a predetermined time period, it enters sleep mode. When the electronic device is in the low-power threshold detection mode, the low-power threshold detection circuit operates to detect the acceleration of the electronic device and determine whether the detected acceleration exceeds a third predetermined threshold range. If not, the low-power threshold detection mode continues; if so, the continuous operating mode is entered. Compared with the prior art, this invention can accurately and timely wake the electronic device from sleep mode, thereby reducing power consumption.

A method performed by a provider computing system includes training one or more artificial intelligence (AI) models to generate task prompts, receiving a requested action from a customer, determining contextual information about a plurality of customer devices, generating a task prompt using the one or more AI models, generating an interaction board accessible by the customer via a customer device and accessible by an employee via an employee device, receiving an indication that the customer pre-task has been completed prior to a customer interaction at a branch location, updating the interaction board to indicate that the customer pre-task has been completed prior to the customer interaction, receiving information associated with the requested action from the customer device during the customer interaction, and updating the interaction board to include the information received from the customer device during the customer interaction.