Composing rich content messages

A computerized method of generating content for an electronic communication, the computerized method comprising:

• detecting a task-intent from user-input content within the electronic communication in a client application, the task-intent being associated with a task for enriching the electronic communication with supplemental content, the task-intent being detected utilizing a first model within a set of machine-learning models configured to detect a plurality of task intents, each model within the set of machine-learning models configured to detect a single task-intent that is different than the task-intents determined by other models within the set of machine-learning models;
• determining a set of candidate task-queries associated with the task-intent detected, at least a subset of candidate task-queries each having at least one field for an entity type associated with the task-intent;
• determining an entity of the entity type for at least one candidate task-query within the set of candidate task-queries based on the user-input content;
• generating at least one task-query suggestion based on the subset of candidate task-queries and the determined entity, each task-query suggestion being associated with a task for enriching content of the electronic communication;
• receiving an indication of a selected task-query selected by a user from the at least one task-query suggestion;
• generating a set of selectable search results for the selected task-query, wherein each selectable search result generated includes supplemental content for the electronic communication that is responsive to the selected task-query and is formatted to be usable by the client application;
• and
• automatically adding the supplemental content to the electronic communication based on a selection of a selectable search result of the selected task-query.

A computerized method for generating a graphical user interface for a computing device, the method comprising:

• observing user content input into an electronic communication within a client application;
• detecting a task-intent from the user content input, the task-intent being associated with a task for enriching the electronic communication with supplemental content, the task-intent being detected utilizing a first model within a set of machine-learning models configured to detect a plurality of task intents, each model within the set of machine-learning models configured to detect a single task-intent that is different than the task-intents determined by other models within the set of machine-learning models;
• determining a set of candidate task-queries associated with the task-intent detected, at least a subset of candidate task-queries each having at least one field for an entity type associated with the task-intent;
• determining an entity of the entity type for at least one candidate task-query within the set of candidate task-queries based on the user content input;
• generating at least one task-query suggestion based on the subset of candidate task-queries and the determined entity, each task-query suggestion being associated with a task for enriching the electronic communication with supplemental content;
• receiving, from a user, a selection of a first task-query from the at least one task-query suggestion;
• generating a set of selectable search results for the first task-query, each selectable search result comprising an object of supplemental content that is responsive to the first task-query;
• and
• automatically initiating a conversational assistant user interface for creating a dialog between a conversational assistant and the user to complete the task associated with the selected first task-query.