Abstract: Techniques are disclosed for automatically generating prompts. A method comprises accessing first prompts, wherein each of the first prompts is a prompt for generating a portion of a SOAP note using a machine-learning model. For each respective first prompt of the first prompts: (i) using the respective first prompt to obtain a first result from a first machine-learning model, (ii) using the respective first prompt and the first result to obtain a second result from a second machine-learning model, the second result including an assessment of the first result, (iii) using the second result to obtain a third result from a third machine-learning model, the third result including a second prompt, (iv) setting the second prompt as the respective first prompt, (v) repeating steps (i)-(iv) a number of times to obtain a production prompt, (vi) adding the production prompt to a collection of prompts; and storing the collection of prompts.

Abstract: Techniques are disclosed for storage and retrieval mechanisms for knowledge artifacts acquired and applicable across conversations to enrich user interactions with a digital assistant. In one aspect, a method includes receiving a natural language utterance form a user during a session between the user and the digital assistant and obtaining a topic context instance for the utterance. The obtaining includes executing a search, determining whether the utterance satisfies a threshold of similarity with one or more topics, identifying the topic context instance associated with the topics, and associating the utterance with the topic context instance. A first generative artificial intelligence model can then be used to generate a list of executable actions. An execution plan is then created, and the topic context instances is updated with the execution plan. The execution plan is then executed, and an output or communication derived from the output is sent to the user.

Abstract: Techniques for time-bound hyperparameter tuning are disclosed. The techniques enable the determination of optimized hyperparameters for a machine learning (ML) model given a specified time bound using a three-stage approach. A series of trials are executed, during each of which the ML model is trained using a distinct set of hyperparameters. In the first stage, a small number of trials are executed to initialize the algorithm. In the second and third stages, a certain number of trials are executed in each stage. The number of trials to run in each stage are determined using one or more computer-implemented techniques. The computer-implemented techniques can also be used to narrow the hyperparameter search space and the feature space. Following the third stage, a set of optimized hyperparameters is adopted based a predefined optimization criterion like minimization of an error function.

Abstract: Techniques are described for performing packet level data centric protection enforcement. Instead of being restricted to perimeter-based security and defining and creating rules that are difficult to maintain, techniques described herein allow users to create data-centric, intent-based policies that are enforced at different enforcement points within one or more networks. In some examples, a method comprises receiving a packet at an enforcement point (EP) within one or more networks that include a plurality of enforcement points (EPs); accessing enforcement data that indicates allowed communications between the EP and one or more other EPs, wherein the data are generated from a policy that specifies how traffic flows the one or more networks and a determination of possible data movements between at least two of EPs in the plurality of EPs; and enforcing the flow of the packet at the EP based on the data.

Abstract: Techniques for managing secure virtual card number (VCN) transactions are disclosed. A POS terminal that processes payments receives an instruction in a secure digital communication over a network to process a payment from a customer to a supplier. Based on receiving a payment request via a network, the POS terminal identifies a VCN associated with the request. The POS terminal validates the VCN and processes the payment request. The POS terminal communicates the VCN to the supplier's bank to initiate a funds transfer between the supplier's bank and the customer's bank that issued the VCN. Upon completion of the transaction, the banks confirm the transaction to the customer and the POS terminal.

Abstract: A summary generation system is disclosed that is configured to generate a summary for content to be summarized by identifying relevant chunks of information from the content to be summarized using a large language model (LLM) and a set of questions. The set of questions enable the system to identify and retrieve relevant chunks of information. Each question undergoes a translation or transformation process to generate multiple question variants for each question. The multiple question variants are used by the system to optimize the search to obtain relevant chunks of information. Then, using the multiple question variants and an LLM, the system extracts information (i.e., answers) from the relevant chunks of information. The summary generation system then collates the answers to create an accurate and comprehensive summary for the content to be summarized.

Abstract: Techniques are disclosed herein for managing ambiguous date mentions in natural language utterances in transforming natural language utterances to logical forms by encoding the uncertainties of the ambiguous date mentions and including the encoded uncertainties in the logical forms. In a training phase, training examples including natural language utterances, logical forms, and database schema information are automatically augmented and used to train a machine learning model to convert natural language utterances to logical form. In an inference phase, input database schema information is augmented and used by the trained machine learning model to convert an input natural language utterance to logical form.

Abstract: Techniques are disclosed herein for implementing digital assistants using generative artificial intelligence. An input prompt comprising a natural language utterance and candidate agents and associated actions can be constructed. An execution plan can be generated using a first generative artificial model based on the input prompt. The execution plan can be executed to perform actions included in the execution plan using agents indicated by the execution plan. A response to the natural language utterance can be generated by a second generative artificial intelligence model using one or more outputs from executing the execution plan.

Abstract: Techniques are disclosed herein for improving the performance of an end-to-end (E2E) Automatic Speech Recognition (ASR) model in a target domain. A set of test examples are generated. The set of test examples comprise multiple subsets of test examples and each subset of test examples corresponds to a particular test category. A machine language model is then used to convert audio samples of the subset of test examples to text transcripts. A word error rate is determined for the subset of test examples. A test category is then selected based on the word error rates and a set of training examples is generated for training the ASR model in a particular target domain from a selected subset of test examples The training examples are used to fine-tune the model in the target domain. The trained model is then deployed in a cloud infrastructure of a cloud service provider.

Abstract: Techniques are disclosed for assisting healthcare providers with common clinical tasks by way of a clinical software application that can be installed on and utilized from various client computing devices. The clinical software application(s) can enable a healthcare provider to record conversations with patients, dictate in natural language, generate patient notes, populate patient records, schedule tasks and generate task notifications, and perform numerous other clinical functions. Applications executing on a mobile computing device and a desktop computing device and concurrently associated with a same user session with a cloud service provider platform, can be paired with one another so that the mobile client application and the desktop client application can operate in concert, under the control of the cloud service provider platform, to provide an end user with a single seamless experience when the end user switches between client devices while performing a task.

Abstract: Techniques are disclosed for automatically generating Subjective, Objective, Assessment and Plan (SOAP) notes. Particularly, techniques are disclosed for automatic SOAP note generation using task decomposition. A text transcript is accessed and segmented into portions. The text transcript can correspond to an interaction between a first entity and a second entity. Machine-learning model prompts are used to extract entities and facts for the respective portions and generate SOAP note sections based at least in-part on the facts. A SOAP note is generated by combining the SOAP note sections. The SOAP note can be stored in a database in association with at least one of the first entity and the second entity.

Abstract: Techniques for using a LLM to detect OOS and OOD utterances. In one aspect, a method includes routing an utterance to a skill bot. The skill bot is configured to execute an action for completing a task associated with the utterance, and a workflow associated with the action includes a GenAI component state configured to facilitate completion of at least part of the task. The method further includes inputting a prompt into a GenAI model for processing. The prompt includes the utterance and scope-related elements that teach the GenAI model to output an invalid input variable when the utterance is OOS or OOD. When the GenAI model determines the utterance is OOS or OOD as part of the processing, the response is generated to include the invalid input variable, and the GenAI component state is caused to transition to a different state or workflow based on the response.

Abstract: Techniques are disclosed for automatically generating Subjective, Objective, Assessment and Plan (SOAP) notes. Particularly, techniques are disclosed for identifying entities for automatic SOAP note generation. A text transcript is accessed and segmented into portions. The text transcript can correspond to an interaction between a first entity and a second entity. Entities for the respective portions are identified using machine-learning models. A SOAP note is generated using the one or more machine-learning models and facts are derived from the text transcript based at least in-part on the entities. The SOAP note can be stored in a database in association with at least one of the first entity and the second entity.

Abstract: Techniques for enhanced chatbot interaction using various large language model providers are provided. In one aspect, a method may include generating a request payload having a common request body specification based on an utterance such that the common request body specification may be a standardized data input format used by a generative artificial intelligence (GenAI) interface for interacting with GenAI model providers. In various embodiments, the method may include converting the common request body specification into a custom request body specification having a data input format associated with a GenAI model provider selected from the plurality of GenAI model providers, communicating, by the GenAI interface, the request payload with the custom request body specification to the GenAI provider for processing by a GenAI model, receiving, at the GenAI interface from the GenAI model provider, a response payload associated with: (i) an error, (ii) processing the request payload, or (iii) both.

Abstract: Techniques are disclosed for assisting healthcare providers with common clinical tasks by way of a clinical software application that can be installed on and utilized from various client computing devices. The clinical software application(s) can enable a healthcare provider to record conversations with patients, dictate in natural language, generate patient notes, populate patient records, schedule tasks and generate task notifications, and perform numerous other clinical functions. A state of the application executing on the client computing devices can be centrally and remotely controlled by a cloud service provider platform. When a user is logged in to both a mobile client computing device and a desktop client computing device, a state of both applications can be concurrently controlled by the cloud service provider platform, and the applications can be linked and synchronized to provide the end user with a seamless experience when moving between the applications.

Abstract: Techniques are disclosed herein for managing date-time intervals in transforming natural language utterances to logical forms by providing an enhanced grammar, a natural language utterance comprising a date-time interval, and database schema information to a machine learning model that has been trained to convert natural language utterances to logical forms; and using the machine learning model to convert the natural language utterance to an output logical form, wherein the output logical form comprises at least one of the date-time interval and an extraction function for extracting date-time information corresponding to the date-time interval from at least one date-time attribute of the database schema information.

Abstract: Techniques for maintaining state and context of conversations between a user and digital assistant using threads. In one aspect, a method includes receiving a natural language utterance from a user during a session, obtaining a topic context instance for the natural language utterance, and generating, by a GenAI model, a list comprising an executable action based on candidate actions associated with the topic context instance. The executable action is then executed to produce an output. The executing includes determining there is no thread running within the session that is associated with the topic context instance, the executable action, or both, and responsive to determining there is no thread running, creating a thread associated with the topic context instance, the executable action, or both, and executing, using the thread, the executable action to obtain the output. The output or a communication derived from the output is then sent to the user.

Abstract: Techniques are disclosed for providing an authenticated model customization for a machine-learning model. A cloud service provider platform accesses a message including, at least, timestamp data and user identification data. A training group of data entities is identified based on the data in the message. A training dataset is determined based on the training group of data entities. A machine-learning model is modified based on the training dataset. The modified machine-learning model is provided during an authenticated network session associated with the user identification data. In some embodiments, the modification of the machine-learning model is removed based on a determination that the authenticated network session had ended.

Abstract: The techniques described herein provide a novel medication order pipeline may be used to facilitate medication orders by identifying the medication ordering intent from a natural language utterance, and using the FHIR-compliance data structure to generate medication order information to fulfill medication orders through an EHR system. The medication order information may be a concise search phrase containing the medical entities extracted from the data structure, or converted EHR system-specific medical codes based on the standard medical codes in the data structure.

Abstract: Techniques are disclosed for fine-tuning a pre-trained machine learning model to be used by a digital assistant for supporting a user's interactions. In one aspect, a method includes accessing a set of training examples, generating a set of synthesized training examples using an iterative process including accessing a dialog script and corresponding prompt template and response template for a predefined scenario, generating one or more prompts based on the dialog script and corresponding prompt template, generating one or more responses associated with each of the one or more prompts based on the dialog script and the response template, and linking each of the responses with the associated prompts to generate one or more synthesized training examples in the set of synthesized training examples. The pre-trained machine learning model is then fine-tuned using the set of training examples and the set of synthesized training examples.