Notion Labs, Inc.

A technique is disclosed for presenting blocks (e.g., pages) that are considered relevant to a user of a workspace, where the pages that are chosen are determined based on the actions of other users whose actions closely align with the user's. Pairs of users in a workspace are given a similarity score based on how likely they are to be interested in the same content of the workspace. This can involve editing the same content, reacting to the same content, or other qualities that indicate a close association. Content is then presented to the user based on the actions of users while factoring in similarity scores.

Disclosed herein are techniques for artificial intelligence (AI) driven conflict detection in collaborative content management. A computing platform having a block-based data structure can generate and display a graphical user interface (GUI) that can include a conflict check entry point. Upon detecting a user interaction with the entry point, the platform can determine and use a block identifier and/or additional contextual information to generate a set of questions regarding content of the block displayed at the GUI. The questions can be generated using intent assessment techniques. Using the generated questions, the platform can query cross-validation compute resources, generate or obtain cross-validation data items, and perform cross-validation of one or more items referenced by the block identifier. Cross-validation operations can be based on content, derived content, metadata, sampling, data schema definitions, other suitable techniques, and/or combinations thereof. The platform can raise alerts and/or facilitate automatic actions in response to cross-validation results.

The present disclosure relates to systems and methods for improved searching. In some implementations, the approaches herein can be used to rank search results to provide more relevant results to users, for example by monitoring the interactions of one or more users with pages or other items. In some implementations, the approaches herein can be used to refine and improve the performance of a search result ranking model over time.

The disclosed technology pertains to a block-based system for creating and organizing documents. The technology includes a method for receiving a user’s indication to create a document and providing guidance through multiple templates selected based on the user’s profile and recent activities. The system allows users to preview templates by creating temporary blocks and instantiating them into permanent blocks upon selection. The block model supports dynamic units of information that can be transformed, moved, and nested within workspaces, enabling flexible customization and organization. The system aims to simplify document creation by offering relevant templates and visualizations, enhancing user experience without extensive training.

A system for providing suggested edits for a block-based workspace displays content organized into multiple blocks on a workspace. Each block is associated with one or more content items. The system can enter a suggestions mode that allows users to input suggested edits or comments. While in the suggestion mode, the system can receive a suggested edit to a content item within a block, presenting it as an inline or block-level edit. The system can generate a suggestion card with a description of the edit in response to the user's input. The system can store an identifier of the suggested edit as a property of the block and display the suggested edit and suggestion card alongside the content item on the page.

A system is configured to generate, by an artificial intelligence (AI) system, AI-suggested edits to content on a page of a workspace. The system can render a preview that integrates the AI-suggested edits into the content as non-persistent suggested edits, which can be individually or collectively accepted or rejected by a user of the workspace. In response to an indication that a first AI-suggested edit to a first content item of a first block is accepted by the user, the system can convert the first AI-suggested edit into a suggested edit and store an indication of the suggested edit as a property of the first block, including the first content item and the suggested edit. In response to an indication that a second AI-suggested edit is rejected by the user, the system can discard the second suggested edit from the preview. The discarded AI-suggested edits can be irreversible.

Methods and devices for applying configurations across all the workspaces that belong to an organization without structurally consolidating the workspaces are disclosed. In one example aspect, an enterprise system includes a first workspace created by a first set of users, a second workspace created a second set of users, and an organization-level space associated with the first workspace and the second workspace. A setting configured in the organization-level space is applied to the first workspace and the second workspace without impacting independent operations of the first workspace and the second workspace.

Methods and devices for applying configurations across all the workspaces that belong to an organization without structurally consolidating the workspaces are disclosed. In one example aspect, an enterprise system includes a first workspace created by a first set of users, a second workspace created a second set of users, and an organization-level space associated with the first workspace and the second workspace. A setting configured in the organization-level space is applied to the first workspace and the second workspace without impacting independent operations of the first workspace and the second workspace.

Described herein are systems, methods, and devices that can be used to aggregate user-specific content across multiple databases. In some implementations, databases within a specified scope are included in the aggregated view. The databases can comprise data used to populate one or more blocks of an integrated workspace, each block comprises one or more properties. The one or more properties can comprise at least one common property. The scope can be a grouping of hierarchically organized blocks of a render tree. The user-specific content can include, for example, tasks, projects, and/or wiki pages.

The disclosed system receives an indication to create tabs in a GUI and creates a hierarchical data structure including a root node and a first child node. The hierarchical data structure represents the tabs having multiple permissions, the root node represents the tabs, and the first child node associated with the root node represents a tab among the tabs. The system creates the tabs that are visible without scrolling on a mobile device screen. The system receives an indication of content to include in the tab and creates a second child node including the content by formatting the content to be visible without scrolling on the mobile device screen. The first child node is a parent of the second child node including the content. The system provides the tabs and the content in the GUI, where viewing the content on the mobile device screen does not involve scrolling.