Intelligent cohorts for network content delivery

1 . A non-transitory machine-readable medium comprising instructions that, when executed by a processor, cause the processor to:

intercept a data packet in transit between a user device and a host through an internet service provider (ISP) core network;

assign the data packet to a category of content;

analyze the data packet to obtain a user identifier associated with the data packet, wherein the user identifier uniquely identifies a user of the user device;

update an identity graph of the user based on the category of content, wherein the identity graph tracks accesses to different categories of content by the user device; and

create or update a cohort of different users using identity graphs of the different users, wherein the cohort comprises a grouping of categories of content with percentage weights assigned to the categories that reflect a general group of users based on usage behavior and belongs to a set of cohorts that are associated with user identifiers, the set of cohorts being configured to be queried using a particular user identifier to return a particular cohort corresponding to the particular user identifier,

wherein the set of cohorts enables determining the particular cohort to which the particular user identifier belongs to enable delivery of content associated with the particular cohort to a particular user represented by the particular user identifier.

2 . The non-transitory machine-readable medium of claim 1 , wherein the instructions are to update the cohort by applying the identity graphs to an artificial intelligence system.

3 . The non-transitory machine-readable medium of claim 1 , wherein the instructions are to create the cohort by applying the identity graphs to an artificial intelligence system.

4 . The non-transitory machine-readable medium of claim 1 , wherein the instructions are to:

determine outcomes of communications between the different user devices and different hosts; and

update the cohort or create a new cohort by applying the identity graphs and the outcomes to an artificial intelligence system.

5 . The non-transitory machine-readable medium of claim 4 , wherein the outcomes include one or more of:

a sale of a good or service;

following a hyperlink;

engaging with content;

viewing an advertisement;

creating or renewing a subscription; and

creating or logging into an account.

6 . The non-transitory machine-readable medium of claim 1 , wherein the instructions are to obtain the user identifier of the data packet by:

extracting the user identifier from the data packet; or

selecting the user identifier from a set of user identifiers using identifying information in the data packet.

7 . The non-transitory machine-readable medium of claim 6 , wherein the identifying information includes one or more of:

an internet protocol (IP) address;

a third-party user identifier;

a hardware identifier of the user device; and

a fingerprint of the user device.

8 . The non-transitory machine-readable medium of claim 1 , wherein the instructions are to assign the data packet to the category of content by categorizing content at a host that is associated with the packet.

9 . A computing device comprising the non-transitory machine-readable medium of claim 1 and a processor to execute the instructions.

10 . A method comprising:

intercepting a data packet in transit between a user device and a host through an internet service provider (ISP) core network;

assigning the data packet to a category of content;

analyzing the data packet to obtain a user identifier associated with the data packet, wherein the user identifier uniquely identifies a user of the user device;

updating an identity graph of the user based on the category of content, wherein the identity graph tracks accesses to different categories of content by the user device; and

creating or updating a cohort of different users using identity graphs of the different users, wherein the cohort comprises a grouping of categories of content with percentage weights assigned to the categories that reflect a general group of users based on usage behavior and belongs to a set of cohorts that are associated with user identifiers, the set of cohorts being configured to be queried using a particular user identifier to return a particular cohort corresponding to the particular user identifier,

wherein the set of cohorts enables determining the particular cohort to which the particular user identifier belongs to enable delivery of content associated with the particular cohort to a particular user represented by the particular user identifier.

11 . The method of claim 10 , further comprising updating the cohort by applying the identity graphs to an artificial intelligence system.

12 . The method of claim 10 , further comprising creating the cohort by applying the identity graphs to an artificial intelligence system.

13 . The method of claim 10 , further comprising:

determining outcomes of communications between the different user devices and different hosts; and

updating the cohort or creating a new cohort by applying the identity graphs and the outcomes to an artificial intelligence system.

14 . The method of claim 10 , further comprising obtaining the user identifier of the data packet by:

extracting the user identifier from the data packet; or

selecting the user identifier from a set of user identifiers using identifying information in the data packet.

15 . A computing device configured to perform the method of claim 10 .