Dropbox, Inc.

Tsung-Hsiang Chang

John Robert Grace, Jr

Igor Kofman

Jacob Hurwitz

Jennifer Lee

David Stinnette

Kavitha Radhakrishnan

Ben Newhouse

Kenneth Choi

Leona Dondi

Susan Wang

Aj Palkovic

Geoff Hulten

Neil Sethi

Anthony Penta

Ashok Pancily Poothiyot

Tim Gasser

A J Palkovic

Ali Zafar

Stephen Voorhees

Ryhan Hassan

Christina Cacioppo

Eric Shobe

Phillip Cohen

The present disclosure relates to systems, non-transitory computer-readable media, and methods for selecting machine-learning models and hardware environments for executing a task. In particular, in one or more embodiments, the disclosed systems select a designated machine-learning model for executing a task based on workload features of the task and task routing metrics for a plurality of machine-learning models. In addition, in one or more embodiments, the disclosed systems select a designated hardware environment for executing the task based on workload features for the task and task routing metrics for a plurality of hardware environments. In some embodiments, the disclosed systems select a fallback machine-learning model and a fallback hardware environment for executing the task if the designated machine-learning model or designated hardware environment are unavailable. Moreover, in one or more embodiments, the disclosed systems can pause and initiate tasks based on bandwidth availability.

Dynamically selecting artificial intelligence models and hardware environments to execute tasks

This disclosure describes systems that identify one or more models (e.g., large language models and/or virtual assistants) permitted to access content items stored for user accounts within a content management system. The disclosed systems can determine a model available to a user account within the content management system from among the one or more models. For example, the disclosed systems can determine one or more relationships between the user accounts within the content management system, large language models utilized by the user accounts, virtual assistants utilized by the user accounts, and content items accessed by the user accounts. The disclosed systems can determine the model for the user account according to the one or more relationships. The disclosed systems can provide a notification corresponding to the model via a user interface of a client device associated with the user account.

Providing contextualized large language model recommendations

Disclosed are systems, methods, and non-transitory computer-readable storage media for managing projects using references between the project and project items. Project items can be, for example, synchronized content items, collaborative content items, other projects, folders, tasks, user accounts, etc. The content management system can create a project identifier for managing data and/or people associated with a project. In various implementations, the content management system can store references between the project and project items in one or more folders associated with the project, in a database, in content item metadata, etc. In some implementations, the storage location of a content item does not affect whether it is associated with the project. When a project is selected by a user, the content management system can generate project view that presents various project items associated with the project in a convenient and easy to access graphical user interface.

Presenting project data managed by a content management system

Disclosed are systems, methods, and non-transitory computer-readable storage media for managing project tasks using content items. A user can create tasks within a content item. The tasks can be assigned to users and/or be assigned due dates. The content item can be shared with other users such that updates to the content item (e.g., updates to the tasks) can cause sending notifications to users associated with the content item. When a task in the content item is updated, all members of the project can be notified of the updated status of the task. A task list for a project can be created that aggregates tasks from multiple content items associated with the project. Tasks can be organized in the task list according to the content item the task was created in. Updates to tasks in the task list can cause corresponding updates to the tasks in the content items.

Managing project tasks using content items

Automatically suggesting project affiliations

A computer-implemented method, system, and non-transitory computer-readable medium may provide event attendees of a network-accessible calendar with a way to generate collaborative notes for a meeting related to an event. In some embodiments, a calendar entry in a network-accessible calendar may be identified. The calendar entry may correspond to an event and may identify the event attendees. A collaborative document that capable of receiving modifications to the event from the event attendees may be created. The event attendees may further be provided access to the collaborative document.

Event-related document generation

Creating projects in a content management system

Managing projects using references

Techniques for optimizing search result snippet selection. In one aspect, the techniques include receiving a query and identifying a set of documents matching the query. A set of one or more snippet field types for a snippet representing a corresponding document of the set of documents is selected based at least in part on a snippet selection model. The snippet selection model is adjusted to reinforce or discourage selection of the set of one or more snippet field types for future queries based on the user click.

Optimizing search result snippet selection

An automated system and method for presenting trending content items to a user in a graphical user interface. The system and method generates trend scores of content items for a user based on current interactions with the content items by top collaborators of the user. By accounting for trends in the form of current content item interactions by collaborators, the system and method determines which content items are both currently relevant to a given user and currently trending among the user's collaborators. The system and method presents content items determined to be trending for a user in a graphical user interface presented to the user. The system and method system improve computers for information discovery by enabling users to spend less time searching for content items.

Growing community of inventors

San Bruno, CA, United States of America

Tsung-Hsiang Chang