Abstract: Systems and methods for using large language models (LLM) to generate and edit content items are described. The methods generate a template of a content item based on user input and populate the template based on data obtained from querying an LLM-generated semantic graph that includes data sources to which a user has authorized access. Once generated, one or more sources provided are used for editing the identified text of the generated content item. If multiple sources are provided, they are weighted based on factors and their priority is determined. Raw data from the provided source(s) are indexed. An LLM using the identified text as an input, queries the index to identify data relevant for editing the identified text. The relevant data is then used to edit, rewrite and/or regenerate the identified text and make the unedited text in the content item coherent with the edited text.

Abstract: The present disclosure is directed to systems and methods of providing content. A server can generate a request for a push content item for an account identifier linked with a computing device. The server can establish a push auction for the account identifier with multiple candidate push content items. The server can determine an auction score for each candidate push content item and select a push content item therefrom based on the auction score. The server can determine a parameter for the account identifier and control delivery of the selected push content item based on a delivery control policy. The server can compare a value of the parameter with a threshold value to authorize the push content item. The server can provide the selected and authorized push content item for presentation in a push content slot via the computing device linked to the account identifier.

Abstract: Systems and methods described receive user input or query at a user interface associated with a large language model (LLM) to perform a task. The user intent and persona are determined based on the user input and other user related data. An LLM automatically generates a workflow consisting of a plurality of steps for performing the task received. Each step of the workflow is mapped to a building block that is to be used for processing the step of the workflow. Selection of the building blocks is based on a plurality of factors, including relevancy, complexity, cost, and accuracy. One of the building blocks used is to perform an application programming interface (API) call to one or more external applications for executing a step of the workflow. Parameters for the API call may be obtained from a generated catalog. Results from the workflow may be customized based on the persona.

Abstract: Systems and methods described receive user input or query at a user interface associated with a large language model (LLM) to perform a task. The user intent and persona are determined based on the user input and other user related data. An LLM automatically generates a workflow consisting of a plurality of steps for performing the task received. Each step of the workflow is mapped to a building block that is to be used for processing the step of the workflow. Selection of the building blocks is based on a plurality of factors, including relevancy, complexity, cost, and accuracy. One of the building blocks used is to perform an application programming interface (API) call to one or more external applications for executing a step of the workflow. Parameters for the API call may be obtained from a generated catalog. Results from the workflow may be customized based on the persona.

Abstract: Systems and methods for using large language models (LLMs) to analyze category specific workflows for generating an optimized workflow that can be used in live conversations to provide a response are described. The methods include categorizing a plurality of projects into specific categories. The actions performed by agents for the plurality of projects are translated into workflows. The workflows are analyzed based on optimization factors and clustering options, such as including redundancies in workflow steps, using alternative solutions to a workflow step, determining whether any escalation performed is justified and if so, adopting escalation related steps. The workflows are consolidated and optimized into an optimized workflow that is tested and verified, and then used in a live conversation.

Abstract: Systems and methods for using large language models (LLM) for generating personalized business content items, such as documents and presentations, by generating content item templates having sections and automatically populating content in its sections by leveraging LLM generated semantic graphs are described. A request to generate a content item is received. A table of contents template having a plurality of nested sections is generated based on the received request. Leveraging an LLM, a semantic graph that semantically organized data from a plurality of enterprise private data sources is generated. A query to the semantic graph for the most fine-gained section of the template is made to obtain relevant indexed data. A write operation is performed to write the data in the most fine-gained section. The query and write operation for each section in the template based on the section's hierarchy is made until all template sections are completed.

Abstract: Systems and methods for using large language models (LLM) to generate and edit content items are described. The methods generate a template of a content item based on user input and populate the template based on data obtained from querying an LLM generated semantic graph that includes data sources to which a user is authorized access. Once generated, on or more sources provided are used for editing the identified text of the generated content item. If multiple sources are provided, they are weighted based on factors and their priority is determined. Raw data from the provided source(s) is indexed. An LLM using the identified text as an input, queries the index to identify data relevant for editing the identified text. The relevant data is then used to edit, rewrite and/or regenerate the identified text and make the other unedited text in the content item coherent with the edited text.

Abstract: Systems and methods for using simple and unstructured conversational input, which is a part of an interactive conversation between a user and a system, to automatically generate a workflow for performing an enterprise task is described. The methods include generating a configurable application that includes the workflow, which when executed, performs the enterprise task. A simple conversational input, which is unstructured data, is processed to determine a task to be performed and a persona that will be performing such a task. Deep learning techniques are applied to leverage related data in the same domain from a knowledge base, that is associated with a foundational transformer model. A workflow involving a plurality of steps is automatically generated in-real time, based on the leveraged learning, while the interactive conversation is still in progress. A user interface that displays both the generated workflow and the input conversation is generated.

Abstract: Systems and methods for training and using large language models to respond to queries and cascading to new models when their performance falls below a threshold during the response generation cycle are described. The methods involve receiving and analyzing a query using heuristics to determine the query's categories and a level of granularity at which its response is to be evaluated. Selecting a language model based on the query analysis and during the generation of the response, evaluating the model's performance for its ability to predict the next segment in a response. The methods score the evaluation and use the score to determine whether the language model's performance exceeds a confidence threshold. If it does not, then during the response is being generated, e.g., at a point before the response generation is completed, cascading from the model to a different model that can perform at a higher confidence level.

Abstract: Systems and methods for using enterprise data to train and generate enterprise related large language models (ELLMs) and selectively combining the generated ELLMs and publicly available LLMs to generate a customized LLM for automatically answering queries are disclosed. Enterprise data is obtained, classified, and curated to generate training data that is used to generate a plurality of contextually distinguished ELLMs. The ELLMs differ based on their relation to various enterprise functions, departments, and level of authorized access to enterprise confidential data. When a question is received from a user, both the identity of the user and context of the question are used to select one or more ELLMs and LLMs for processing the question. The processing involves sequential selection of a plurality of ELLMs to process the question, blend the results from the processing, and use the results as input into an ensemble model to obtain a final answer.

Abstract: The present disclosure is directed to systems and methods of providing content. A server can generate a request for a push content item for an account identifier linked with a computing device. The server can establish a push auction for the account identifier with multiple candidate push content items. The server can determine an auction score for each candidate push content item and select a push content item therefrom based on the auction score. The server can determine a parameter for the account identifier and control delivery of the selected push content item based on a delivery control policy. The server can compare a value of the parameter with a threshold value to authorize the push content item. The server can provide the selected and authorized push content item for presentation in a push content slot via the computing device linked to the account identifier.

Abstract: A method includes receiving, by one or more processors of at least one server and via a connection between the at least one server and a client device, a request for a content item to be presented at the client device, receiving, by the one or more processors, data indicative of a network speed of the client device at which the content item is to be presented, selecting, by the one or more processors and based on the data indicative of the network speed, a first content item, and providing, by the one or more processors and for presentation at the client device, the first content item via the connection between the at least one server and the client device.

Abstract: The present disclosure is directed to systems and methods of providing content. A server can generate a request for a push content item for an account identifier linked with a computing device. The server can establish a push auction for the account identifier with multiple candidate push content items. The server can determine an auction score for each candidate push content item and select a push content item therefrom based on the auction score. The server can determine a parameter for the account identifier and control delivery of the selected push content item based on a delivery control policy. The server can compare a value of the parameter with a threshold value to authorize the push content item. The server can provide the selected and authorized push content item for presentation in a push content slot via the computing device linked to the account identifier.

Abstract: Selection and serving of content items may include receiving data indicative of a status of an energy source of a device with a request for a content item. A first received content item may be associated with a first energy consumption level and a second received content item may be associated with a second energy consumption level. The accessed content items are responsive to the request for a content item. The first energy consumption level may be higher than the second energy consumption level. The first content item or the second content item may be selected based, at least in part, on the received data indicative of the status of the energy source of the device, and data to display the selected content item may be provided to the device.

Abstract: In a method, performed by a server, for presenting content on a user device during load latency, a content item for presentation on the user device is selected. The content item and intervening content are transmitted to the user device. The intervening content includes a static version of a landing page associated with the content item. A request to load the landing page is received from the user device. The request is initiated by selection of the content item on the user device. The transmitting occurs before receiving any request to load the landing page from the user device.

Abstract: A computer-implemented method for annotating online content with offline interaction data and offline interaction conversion data is implemented by a content annotating computing device in communication with a memory. The method includes receiving a set of content interaction data based on an online interaction between an online user and a first online content item, identifying at least one physical location associated with the online content provider, determining a set of offline interaction conversion data based on the set of content interaction data and a set of offline interaction data, receiving a second online content item associated with the online content provider, and annotating the second online content item based at least partially on the set of offline interaction conversion data and the set of offline interaction data.

Abstract: A computer-implemented method for annotating online content with offline interaction data and offline interaction conversion data is implemented by a content annotating computing device in communication with a memory. The method includes receiving a set of content interaction data based on an online interaction between an online user and a first online content item, identifying at least one physical location associated with the online content provider, determining a set of offline interaction conversion data based on the set of content interaction data and a set of offline interaction data, receiving a second online content item associated with the online content provider, and annotating the second online content item based at least partially on the set of offline interaction conversion data and the set of offline interaction data.

Abstract: Selection and serving of content items may include receiving data indicative of a status of an energy source of a device with a request for a content item. A first received content item may be associated with a first energy consumption level and a second received content item may be associated with a second energy consumption level. The accessed content items are responsive to the request for a content item. The first energy consumption level may be higher than the second energy consumption level. The first content item or the second content item may be selected based, at least in part, on the received data indicative of the status of the energy source of the device, and data to display the selected content item may be provided to the device.

Abstract: Selection and serving of content items may include receiving data indicative of a status of an energy source of a device with a request for a content item. A first received content item may be associated with a first energy consumption level and a second received content item may be associated with a second energy consumption level. The accessed content items are responsive to the request for a content item. The first energy consumption level may be higher than the second energy consumption level. The first content item or the second content item may be selected based, at least in part, on the received data indicative of the status of the energy source of the device, and data to display the selected content item may be provided to the device.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for facilitating and improving push messaging provided to users. In one aspect, a method includes receiving, from a user device, a request to subscribe to a push messaging service, selecting a given phone number that is uniquely associated with the user based on the received request, transmitting, to a user device, a first message that presents visual information at the user device and includes the given phone number, receiving, by the one or more servers, a second message addressed to the given phone number, selecting an additional message that presents additional visual information at the user device based, at least in part, on the message received from the user device and one or more of a geolocation of the mobile device or a time of day, and transmitting the additional message to the user device.