Verily Life Sciences LLC

Nikhil Roy

Peilun Shan

Aurora Marian Adkins

Thomas Rudick

Lucy Boyd Schachter

Bella Powers

Aaron Haywood Stoertz

Miguel Angelo Drumond

Katherine K Nadell

David Wright

Sarah Abramson

Jesse Elds

Howard Zisser

Cara Leggett Silver

A computer-implemented method improves efficacy of diabetes self-management to obtain a measurement message and/or contextual activity information used to determine a suitable insulin dosage to treat a patient. The method includes simulating a health coach in a natural conversation to obtain a measurement of the patient's blood-glucose level, processing the measured blood-glucose level with a personalized insulin titration algorithm to determine an insulin dosage recommendation, and generating an electronic signal configured to cause a corresponding action to be performed by a computing device.

Virtual health coach for facilitating dynamic dosage titration

A technique for generating and presenting context-aware recommendations to assist a patient in managing a chronic health condition. A health management platform provides various services that are accessible to a patient via a graphical user interface (GUI). Input data are received from various services and processed to generate contextual data associated with the patient such as a health state of the patient. Using the contextual data, a recommendation to access a particular service is generated and presented to the patient via the GUI. For example, in response to determining a health state of a patient, a recommendation to access a service to improve the health state of the patient can be automatically presented via the GUI to encourage the patient to access the service.

Context-aware recommendations in a health management platform user interface

A user interface for a health management platform is described that includes a virtual sidekick configured to assist a patient in managing a chronic health condition. The virtual sidekick may represent an individual (not necessarily the patient) that is facing similar challenges as the patient such as the same chronic health condition as the patient. In some embodiments, the virtual sidekick can be presented in a virtual scene that includes one or more reverse augmented realty elements that are indicative of a physical context of the patient. For example, the virtual scene may depict visual elements indicative of the patient's physical location and/or weather conditions at the patient's physical location. Data used to personalize and dynamically animate the virtual sidekick to assist the patient in managing a chronic health condition can come from multiple different sources.

Virtual sidekick in a health management platform user interface

A technique is described for performing audio-based patient surveys to obtain patient data for use in a health management platform. A virtual assistant is simulated to participate in a natural language conversation with a patient. As part of the conversation, the virtual assistant generates and outputs an audible natural language message that is then presented via a speaker at a user computing device. The audible natural language message includes a prompt for the user to provide a natural language reply. Reply data indicative of a reply by the user is then received and processed to generate and/or update patient data associated with the patient. The patient data is then applied by the health management platform to assist the patient in managing a chronic health condition.

Audio-based patient surveys in a health management platform

A computer-implemented method for dynamically reallocating resources to users of a coaching service. The method can include initially allocating resources to users of the coaching service, receiving indications of activities of the users relative to their allocated resources, and comparing the activities relative to a reference value to produce an outcome including a likelihood that a target user will benefit from a target resource. Upon identifying the target user, a dynamic reallocation process deallocates the target resource from another user and reallocates the target resource to the target user.

Dynamic and targeted allocation of resources for coaching service

Virtual health coach based insulin titration platform

A computer-implemented method of a coaching service with variable content objects. The method can include invoking a content flow including an ordered sequence of content objects. The content flow is arranged in accordance with a coaching protocol for a user of a user device. The user accesses a coaching service through the user device. The method further includes causing the user device to present a first content object. The first content object is prearranged in the ordered sequence of content objects based on information associated with the user. The method further includes dynamically selecting a second content object to replace a next content object of the ordered sequence of content objects and advancing the content flow to the second content object in lieu of the next content object and in accordance with the coaching protocol.

Variable content customization for coaching service

Introduced here are diabetes management platforms able to guide people with diabetes toward a glycemic target. Rather than state the absolute amount of glucose within the blood, the diabetes management platform can instead produce personalized glycemic goals based on the physiological data associated with an individual. For example, if the diabetes management platform determines that the time spent within a first glycemic range exceeds a specified threshold, then the diabetes management platform may recommend that the individual attempt to keep their blood glucose level within a second glycemic range. Generally, the second glycemic range is closer than the first glycemic range to a target glycemic range corresponding to a healthy glycemic state.

Tiered time-in-range guiding interfaces for diabetes management

The disclosed embodiments include a computer-implemented method for improving efficacy of diabetes self-management with automated personalized computer engagement that reliably obtains a measurement message and/or contextual information used to determine a suitable insulin dosage of an insulin titration process to treat the diabetic patient. The method includes causing display of a messaging interface that couples the user of the computing device to a chat bot simulating a health coach in a natural conversation to obtain a measurement of the diabetic patient's blood-glucose level. The method includes determining an insulin dosage recommendation to treat the diabetic patient by processing the measured blood-glucose level with an insulin titration algorithm that is personalized for the diabetic patient, and generating an electronic signal configured to cause a corresponding action to be performed by a computing device (e.g., display message of recommended insulin dosage).

Introduced here are techniques for developing personalized retroactive insights into how contextual factors can affect glycemic responses. The personalized retroactive insights can be made available to the corresponding individual so that they can examine the impact certain contextual events have had on blood glucose level. A contextual event represents an activity or a circumstance that is related to glycemic response. Rather than state the absolute amount of certain molecules (e.g., carbohydrates, protein, or fat) in a foodstuff, an insight detection platform can instead interpret disparate data types to discover the effect certain contextual events have on blood glucose level.

Growing community of inventors

Sunnyvale, CA, United States of America

Nikhil Roy