Machine learning system for predicting optimal interruptions based on biometric data collected using wearable devices

1. A method, comprising:

receiving, at a first moment in time, first biometric data for a plurality of users from a plurality of sensor devices;

identifying, for each respective user of the plurality of users, a respective user category based on characteristics of the respective user;

retrieving, for each respective user category, a respective trained machine learning model, wherein the respective trained machine learning model was trained specifically for users belonging to the respective user category;

generating a group metric by processing the first biometric data using at least one trained machine learning model, wherein biometric data associated with each respective user is processed using the respective trained machine learning model corresponding to the respective user category identified based on characteristics of the respective user, comprising:

processing a first subset of the first biometric data using a first trained machine learning model based on determining that each user associated with the first subset belongs to a first user category for which the first trained machine learning model was trained; and

processing a second subset of the first biometric data using a second trained machine learning model based on determining that each user associated with the second subset belongs to a second user category for which the second trained machine learning model was trained;

determining that the group metric does not satisfy one or more predefined criteria;

receiving, at a second moment in time, second biometric data for the plurality of users from the plurality of sensor devices;

generating an updated group metric by processing the second biometric data using the at least one trained machine learning model; and

upon determining that the updated group metric satisfies the one or more predefined criteria, providing an indication that the one or more predefined criteria have been satisfied.

2. The method of claim 1 , wherein the plurality of users are engaged in a group activity, and wherein the trained machine learning model is further refined using data from the plurality of users.

3. The method of claim 2 , the method further comprising:

upon providing the indication that the one or more predefined criteria have been satisfied, receiving, from at least one of the plurality of users, an indication that the at least one user wants to continue engaging in the group activity; and

updating the at least one trained machine learning model based on the received indication.

4. The method of claim 2 , the method further comprising:

receiving, from at least one of the plurality of users, an indication that the at least one user wants to stop engaging in the group activity, wherein the indication is received prior to determining that the updated group metric satisfies the one or more predefined criteria; and

updating the at least one trained machine learning model based on the received indication.

5. The method of claim 2 , the method further comprising:

determining, after providing the indication that the one or more predefined criteria have been satisfied, that the plurality of users are still engaged in the group activity; and

updating the at least one trained machine learning model based on the determination.

6. The method of claim 2 , the method further comprising:

determining, before the one or more predefined criteria has been satisfied, that the plurality of users have stopped engaging in the group activity; and

updating the at least one trained machine learning model based on the determination.

7. The method of claim 1 , wherein generating the group metric comprises:

aggregating the first biometric data from at least some of the plurality of users; and

processing the aggregated biometric data using the at least one trained machine learning model.

8. The method of claim 1 , wherein generating the group metric comprises:

processing the biometric data for each respective user of the plurality of users using the at least one trained machine learning model to produce a plurality of individual metrics; and

aggregating the individual metrics to form the group metric.

9. The method of claim 1 , wherein the biometric data corresponding to at least one user is processed using a trained machine learning model that was trained using data from the at least one user.

10. The method of claim 1 , wherein the biometric data includes at least one of:

(i) a glucose level from each user;

(ii) a lactate level from each user;

(iii) a sodium level from each user;

(iv) a potassium level from each user;

(v) a heart rate from each user; and

(vi) a temperature of each user.

11. A system, comprising:

a processor; and

a computer memory storing a program, which, when executed on the processor, performs an operation comprising:

receiving, at a first moment in time, first biometric data for a plurality of users from a plurality of sensor devices;

identifying, for each respective user of the plurality of users, a respective user category based on characteristics of the respective user;

retrieving, for each respective user category, a respective trained machine learning model, wherein the respective trained machine learning model was trained specifically for users belonging to the respective user category;

generating a group metric by processing the first biometric data using at least one trained machine learning model, wherein biometric data associated with each respective user is processed using the respective trained machine learning model corresponding to the respective user category identified based on characteristics of the respective user, comprising:

processing a first subset of the first biometric data using a first trained machine learning model based on determining that each user associated with the first subset belongs to a first user category for which the first trained machine learning model was trained; and

processing a second subset of the first biometric data using a second trained machine learning model based on determining that each user associated with the second subset belongs to a second user category for which the second trained machine learning model was trained;

determining that the group metric does not satisfy one or more predefined criteria;

receiving, at a second moment in time, second biometric data for the plurality of users from the plurality of sensor devices;

generating an updated group metric by processing the second biometric data using the at least one trained machine learning model; and

upon determining that the updated group metric satisfies the one or more predefined criteria, providing an indication that the one or more predefined criteria have been satisfied.

12. The system of claim 11 , wherein the plurality of users are engaged in a group activity, and wherein the trained machine learning model is further refined using data from the plurality of users, the operation further comprising:

determining, after providing the indication that the one or more predefined criteria have been satisfied, that the plurality of users are still engaged in the group activity; and

updating the at least one trained machine learning model based on the determination.

13. The system of claim 11 , wherein the plurality of users are engaged in a group activity, and wherein the trained machine learning model is further refined using data from the plurality of users, the operation further comprising:

determining, before the one or more predefined criteria have been satisfied, that the plurality of users have stopped engaging in the group activity; and

updating the at least one trained machine learning model based on the determination.

14. The system of claim 11 , wherein generating the group metric comprises:

aggregating the first biometric data from the plurality of users; and

processing the aggregated biometric data using the at least one trained machine learning model.

15. The system of claim 11 , wherein the biometric data corresponding to at least one user is processed using a trained machine learning model that was trained using data from the at least one user.

16. A computer-readable storage medium having computer-readable program code embodied therewith, the computer-readable program code executable by one or more computer processors to perform an operation comprising:

receiving, at a first moment in time, first biometric data for a plurality of users from a plurality of sensor devices;

identifying, for each respective user of the plurality of users, a respective user category based on characteristics of the respective user;

retrieving, for each respective user category, a respective trained machine learning model, wherein the respective trained machine learning model was trained specifically for users belonging to the respective user category;

generating a group metric by processing the first biometric data using at least one trained machine learning model, wherein biometric data associated with each respective user is processed using the respective trained machine learning model corresponding to the respective user category identified based on characteristics of the respective user, comprising:

processing a first subset of the first biometric data using a first trained machine learning model based on determining that each user associated with the first subset belongs to a first user category for which the first trained machine learning model was trained; and

processing a second subset of the first biometric data using a second trained machine learning model based on determining that each user associated with the second subset belongs to a second user category for which the second trained machine learning model was trained;

determining that the group metric does not satisfy one or more predefined criteria;

receiving, at a second moment in time, second biometric data for the plurality of users from the plurality of sensor devices;

generating an updated group metric by processing the second biometric data using the at least one trained machine learning model; and

upon determining that the updated group metric satisfies the one or more predefined criteria, providing an indication that the one or more predefined criteria have been satisfied.

17. The computer-readable storage medium of claim 16 , wherein the plurality of users are engaged in a group activity, and wherein the trained machine learning model is further refined using data from the plurality of users, the operation further comprising:

determining, after providing the indication that the one or more predefined criteria have been satisfied, that the plurality of users are still engaged in the group activity; and

updating the at least one trained machine learning model based on the determination.

18. The computer-readable storage medium of claim 16 , wherein the plurality of users are engaged in a group activity, and wherein the trained machine learning model is further refined using data from the plurality of users, the operation further comprising:

determining, before the one or more predefined criteria have been satisfied, that the plurality of users have stopped engaging in the group activity; and

updating the at least one trained machine learning model based on the determination.

19. The computer-readable storage medium of claim 16 , wherein generating the group metric comprises:

aggregating the first biometric data from the plurality of users; and

processing the aggregated biometric data using the at least one trained machine learning model.

20. The computer-readable storage medium of claim 16 , wherein the biometric data corresponding to at least one user is processed using a trained machine learning model that was trained using data from the at least one user.