Communication platform

A system for sharing machine-learning-generated insights on a communication platform, comprising:

• one or more processors,
• a memory, and
• one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, the one or more programs including instructions for:
• causing display, within a user interface of the communication platform associated with a user, a portion of a conversation between the user and one or more conversation participants,
• constructing a first plurality of embedding representations of a plurality of messages in the conversation, wherein the first plurality of embedding representations:
• is associated with permissions, wherein the permissions associated with the first plurality of embedding representations are based on one or more ethical wall within an organization, and
• comprises first vector representations generated by providing the plurality of messages to a self-supervised model;
• storing the first plurality of embedding representations in a database;
• providing, from the database, the first plurality of embedding representations to one or more machine-learning models;
• generating, using the one or more machine-learning models and the first plurality of embedding representations, a first insight based on the conversation between the user and the one or more conversation participants;
• causing display, within the user interface of the communication platform associated with the user, the first insight in association with the conversation;
• receiving a user request to share the first insight with a contact on the communication platform, wherein the contact is not the one or more conversation participants;
• after receiving the user request to share the first insight with the contact, determining, according to the permissions associated with the first plurality of embedding representations, whether the contact has permission to access the first insight;
• in accordance with a determination that the contact has the permission to access the first insight, causing display, within a user interface of the communication platform associated with the contact, the first insight without displaying messages of the conversation;
• receiving, via a source different from the conversation, feedback from the user;
• and
• constructing a second plurality of embedding representations of the feedback comprising second vector representations generated by providing the feedback to the self-supervised model, wherein a second insight is generated based on the conversation between the user and the one or more conversation participants, and further based on the second plurality of embedding representations.

A non-transitory computer-readable storage medium storing one or more programs for sharing machine-learning-generated insights on a communication platform, the one or more programs comprising instructions, which when executed by one or more processors of one or more electronic devices, cause the one or more electronic devices to:

• cause display, within a user interface of the communication platform associated with a user, a portion of a conversation between the user and one or more conversation participants;
• construct a first plurality of embedding representations of a plurality of messages in the conversation, wherein the first plurality of embedding representations:
• is associated with permissions, wherein the permissions associated with the first plurality of embedding representations are based on one or more ethical wall within an organization, and
• comprises first vector representations generated by providing the plurality of messages to a self-supervised model;
• store the first plurality of embedding representations in a database;
• provide, from the database, the first plurality of embedding representations to one or more machine-learning models;
• generate, using the one or more machine-learning models and the first plurality of embedding representations, a first insight based on the conversation between the user and the one or more conversation participants;
• cause display, within the user interface of the communication platform associated with the user, the first insight in association with the conversation;
• receive a user request to share the first insight with a contact on the communication platform, wherein the contact is not the one or more conversation participants;
• after receiving the user request to share the first insight with the contact, determine, according to the permissions associated with the first plurality of embedding representations, whether the contact has permission to access the first insight;
• in accordance with a determination that the contact has the permission to access the first insight, cause display, within a user interface of the communication platform associated with the contact, the first insight without displaying messages of the conversation;
• receive, via a source different from the conversation, feedback from the user;
• and
• construct a second plurality of embedding representations of the feedback comprising second vector representations generated by providing the feedback to the self-supervised model, wherein a second insight is generated based on the conversation between the user and the one or more conversation participants, and further based on the second plurality of embedding representations.