Granted Patent B2

US 12,568,059 · App. 18/501,624 · Granted Mar 3, 2026

Inventors: Curtis Neil Allen, Jr. (Denver, CO); David Barrett-Kahn (New York City, NY); Zachary Alan Kaarvik (New Westminster, CA); Omar Lee (Berkeley, CA); McKenna Lowry (Valhalla, NY); Kelly Holmes Moran (Brooklyn, NY); Sohom Paul (Sugar Land, TX); Jacquelyn Elizabeth Rocca (San Francisco, CA); Leonard Jackson Shearer (San Francisco, CA); Katherine Jane Steigman (San Leandro, CA); Wii Yatani (Roslyn, NY)

Assignee: Salesforce, Inc.

H04L51/216H04L51/04H04L51/10H04L51/212

Abstract

Claims (67)

1 . A method, implemented at least in part by one or more computing devices associated with a group-based communication platform, comprising:

receiving, from a first user of the group-based communication platform, a first indication of a message previously sent to a virtual space of the group-based communication platform, wherein the first user is a member of the virtual space;

determining one or more characteristics associated with the virtual space, including at least an engagement characteristic that is greater than a threshold engagement and defined relative to a number of members of the virtual space, wherein the one or more characteristics associated with the virtual space are determined, at least in part, based on messages previously sent to the virtual space by two or more other members of the virtual space different from the first user;

receiving, from the first user, a second indication to draft an intended message to send to the virtual space;

receiving, from the first user, a request to modify the intended message;

inputting the intended message, the request to modify the intended message, and the one or more characteristics associated with the virtual space into a machine-learning model;

receiving, from the machine-learning model, a modified message based at least in part on the intended message and the one or more characteristics associated with the virtual space, wherein the modified message is associated with a confidence score indicating a likelihood that the modified message will be sent by the first user to the virtual space without further modification;

causing presentation, to the first user and via the one or more computing devices, of the modified message;

receiving, from the first user, a third indication of a selection of the modified message to send the modified message to the virtual space; and

sending, based at least in part on the third indication of the selection of the modified message, the modified message to the virtual space, wherein sending the modified message to the virtual space comprises identifying a plurality of user identifiers associated with respective members of the virtual space to receive the modified message.

2 . The method of claim 1 , wherein the engagement characteristic is one of a number of replies to the message previously sent to the virtual space, a number of viewers of the message previously sent to the virtual space, or a number of reactions to the message previously sent to the virtual space, and wherein the one or more characteristics associated with the virtual space further include:

addition of one or more added users to the virtual space;

removal of one or more removed users from the virtual space;

creation of one or more tasks associated with the virtual space;

completion of the one or more tasks associated with the virtual space; or

creation of one or more workflows associated with the virtual space.

3 . The method of claim 1 , wherein the request to modify the intended message includes a sentiment to be associated with the modified message, and wherein the modified message includes language associated with the sentiment.

4 . The method of claim 1 , further comprising:

determining that the intended message contains one or more profanities; and

refraining from generating the modified message.

5 . The method of claim 1 , wherein the modified message is generated using a large language model.

6 . The method of claim 1 , wherein the virtual space is one of a workspace, a communication channel, a direct message instance, a canvas, or a collaborative document.

7 . The method of claim 1 , wherein the modified message is a first modified message having a highest confidence score of a plurality of confidence scores associated with a plurality of modified messages and the method further comprising:

causing presentation, to the first user and via the one or more computing devices associated with the first user, of the first modified message based at least in part on the first modified message having the highest confidence score of the plurality of confidence scores associated with the plurality of modified messages.

8 . A system comprising:

one or more processors; and

one or more non-transitory computer-readable media comprising instructions that, when executed by the one or more processors, cause the system to perform operations comprising:

determining, based at least in part on a plurality of messages previously sent to a virtual space of a group-based communication platform by two or more members of the virtual space, an engagement characteristic of the virtual space, wherein the engagement characteristic is greater than a threshold engagement and defined relative to a number of members of the virtual space;

receiving, from a first user that is a member of the virtual space, a request to modify a message to send to the virtual space;

inputting the request to modify the message and the engagement characteristic of the virtual space into a machine-learning model;

receiving, from the machine-learning model and based at least in part on inputting the request and the engagement characteristic of the virtual space, a modified message, wherein the modified message was generated by the machine-learning model and is associated with a confidence score indicating a likelihood that the modified message will be sent to the virtual space by the first user without further modification; and

causing presentation, to the first user and via one or more computing devices associated with the first user, of the modified message.

9 . The system of claim 8 , the operations further comprising inputting, prior to receiving the modified message, one or more other characteristics associated with the virtual space into the machine-learning model, wherein the one or more other characteristics associated with the virtual space include:

a time period;

addition of one or more added users to the virtual space;

removal of one or more removed users from the virtual space;

creation of one or more tasks associated with the virtual space;

completion of one or more tasks associated with the virtual space; and

creation of one or more workflows associated with the virtual space.

10 . The system of claim 8 , wherein the request to modify the message includes a sentiment to be associated with the modified message, and wherein the modified message includes language associated with the sentiment.

11 . The system of claim 8 , the operations further comprising:

determining that the message to send to the virtual space of the group-based communication platform contains one or more profanities; and

refraining from generating the modified message.

12 . The system of claim 8 , wherein the modified message is generated using a large language model.

13 . The system of claim 8 , wherein the virtual space is one of a workspace, a communication channel, a direct message instance, a canvas, or a collaborative document.

14 . The system of claim 8 , wherein the modified message is a first modified message having a highest confidence score of a plurality of confidence scores associated with a plurality of modified messages and the operations further comprising:

causing presentation, to the first user and via the one or more computing devices associated with the first user, of the first modified message based at least in part on the first modified message having the highest confidence score of the plurality of confidence scores associated with the plurality of modified messages.

15 . One or more non-transitory computer-readable media comprising instructions that, when executed by one or more processors, cause the one or more processors to perform operations comprising:

determine, based at least in part on a plurality of messages previously sent to a virtual space of a group-based communication platform by two or more members of the virtual space, an engagement characteristic of the virtual space, wherein the engagement characteristic is greater than a threshold engagement and defined relative to a number of members of the virtual space;

receive, from a first user that is a member of the virtual space, a request to modify a message to send to the virtual space;

input the request to modify the message and the engagement characteristic of the virtual space into a machine-learning model;

receive, from the machine-learning model and based at least in part on inputting the request and the engagement characteristic of the virtual space, a modified message, wherein the modified message was generated by the machine-learning model and is associated with a confidence score indicating a likelihood that the modified message will be sent to the virtual space by the first user without further modification; and

cause presentation, to the first user and via one or more computing devices associated with the first user, of the modified message.

16 . The one or more non-transitory computer-readable media of claim 15 , the operations further comprising inputting, prior to receiving the modified message, one or more other characteristics associated with the virtual space into the machine-learning model, wherein the one or more other characteristics associated with the virtual space include:

addition of one or more added users to the virtual space;

removal of one or more removed users from the virtual space;

creation of one or more tasks associated with the virtual space;

completion of the one or more tasks associated with the virtual space; and

creation of one or more workflows associated with the virtual space.

17 . The one or more non-transitory computer-readable media of claim 15 , the operations further comprising:

determining that the message to send to the virtual space of the group-based communication platform contains one or more profanities; and

refraining from generating the modified message.

18 . The one or more non-transitory computer-readable media of claim 15 , wherein the modified message is generated using a large language model.

19 . The one or more non-transitory computer-readable media of claim 15 , the operations further comprising:

receiving, from the first user, an indication of a selection of the modified message to send to the virtual space of the group-based communication platform; and

sending, based at least in part on receiving the indication of the selection of the modified message by the first user, the modified message to the virtual space of the group-based communication platform.

20 . The one or more non-transitory computer-readable media of claim 15 , wherein the virtual space is one of a workspace, a communication channel, a direct message instance, a canvas, or a collaborative document.

Assignments (1)

Continuity (1)

References Cited (23)