Contextualizing data to provide computing activity synthesis

1 . A computer-implemented method comprising:

aggregating, for a user account, activity data items having different data formats from one or more data sources, each activity data item comprising contextual content indicating a user interaction within one or more computing environments;

transforming, using a large language model, the activity data items to synthesized activity items by converting the different data formats of the activity data items to the synthesized activity items having a unified textual format;

generating an activity meta summary that describes relationships between the synthesized activity items for a specified timeframe by providing the synthesized activity items to the large language model;

identifying, based on the activity meta summary, a subset of synthesized activity items for the specified timeframe; and

providing, for display on a client device associated with the user account, the subset of synthesized activity items for the specified timeframe.

2 . The computer-implemented method of claim 1 , wherein transforming the activity data items to the synthesized activity items comprises:

generating a synthesizer prompt comprising a prompt instruction to transform the activity data items to the synthesized activity items; and

providing the synthesizer prompt and the activity data items to the large language model.

3 . The computer-implemented method of claim 1 , wherein generating the activity meta summary for the specified timeframe by processing the synthesized activity items with the large language model comprises:

generating a meta prompt comprising a prompt instruction to generate the activity meta summary using the synthesized activity items; and

providing the meta prompt and the synthesized activity items to the large language model.

4 . The computer-implemented method of claim 1 , further comprising:

determining an impact metric to the user account for each of the synthesized activity items; and

identifying the subset of synthesized activity items according to the impact metric.

5 . The computer-implemented method of claim 1 , further comprising:

identifying a second user account;

determining, within the activity data items, collaboration events between the user account and the second user account for the specified timeframe; and

identifying the subset of synthesized activity items according to the collaboration events between the user account and the second user account for the specified timeframe.

6 . The computer-implemented method of claim 1 , further comprising:

identifying an operational initiative;

determining, within the activity data items, operational initiative events for the specified timeframe; and

identifying the subset of synthesized activity items according to the operational initiative events for the specified timeframe.

7 . The computer-implemented method of claim 1 , further comprising:

providing, to the client device associated with the user account, a notification of an activity synopsis; and

based on receiving an indication of a user interaction with the notification, providing the activity synopsis comprising the subset of synthesized activity items for the specified timeframe.

8 . The computer-implemented method of claim 1 , further comprising determining the specified timeframe from the synthesized activity items based on one of:

time stamps of the activity data items;

a subset of the activity data items; or

input from the user account.

9 . A system comprising:

at least one processor; and

a non-transitory computer-readable medium storing instructions which, when executed by the at least one processor, cause the system to:

aggregate, for a user account, activity data items having different data formats from one or more data sources, each activity data item comprising contextual content indicating a user interaction within one or more computing environments;

transform, using a large language model, the activity data items to synthesized activity items by converting the different data formats of the activity data items to the synthesized activity items having a unified textual format;

generate an activity meta summary that describes relationships between the synthesized activity items according to a contextual vector by providing the synthesized activity items to the large language model;

identify, based on the activity meta summary, a subset of synthesized activity items according to the contextual vector; and

provide, for display on a client device associated with the user account, the subset of synthesized activity items according to the contextual vector.

10 . The system of claim 9 , further storing instructions which, when executed by the at least one processor, cause the system to transform the activity data items to the synthesized activity items by:

generating a synthesizer prompt comprising a prompt instruction to transform the activity data items to the synthesized activity items; and

providing the synthesizer prompt and the activity data items to the large language model.

11 . The system of claim 9 , further storing instructions which, when executed by the at least one processor, cause the system to generate the activity meta summary by:

generating a meta prompt comprising a prompt instruction to generate the activity meta summary using the synthesized activity items; and

providing the meta prompt and the synthesized activity items to the large language model.

12 . The system of claim 9 , further storing instructions which, when executed by the at least one processor, cause the system to select the contextual vector from one or more of: a specified timeframe; collaboration events; or operational initiative events.

13 . The system of claim 12 , further storing instructions which, when executed by the at least one processor, cause the system to:

determine an impact metric to the user account for each of the synthesized activity items for the specified timeframe; and

identify the subset of synthesized activity items according to the impact metric.

14 . The system of claim 12 , further storing instructions which, when executed by the at least one processor, cause the system to:

identify a second user account;

determine, within the activity data items, the collaboration events between the user account and the second user account for the specified timeframe; and

identify the subset of synthesized activity items according to the collaboration events between the user account and the second user account for the specified timeframe.

15 . The system of claim 12 , further storing instructions which, when executed by the at least one processor, cause the system to:

determine, within the activity data items, the operational initiative events for the specified timeframe; and

identify the subset of synthesized activity items according to the operational initiative events for the specified timeframe.

16 . A non-transitory computer-readable medium storing executable instructions which, when executed by at least one processor, cause the at least one processor to:

aggregate, for a user account, activity data items having different data formats from one or more data sources, each activity data item comprising contextual content indicating a user interaction within one or more computing environments;

transform, using a large language model, the activity data items to synthesized activity items by converting the different data formats of the activity data items to the synthesized activity items having a unified textual format;

generate an activity meta summary that describes relationships between the synthesized activity items for a specified timeframe by providing the synthesized activity items to the large language model;

generate, based on the activity meta summary, an activity synopsis for the specified timeframe; and

provide, for display on a client device associated with the user account, the activity synopsis for the specified timeframe.

17 . The non-transitory computer-readable medium of claim 16 , further storing instructions which, when executed by the at least one processor, cause the at least one processor to transform the activity data items to synthesized activity items by:

generating a synthesizer prompt comprising a prompt instruction to transform the activity data items to the synthesized activity items; and

providing the synthesizer prompt and the activity data items to the large language model.

18 . The non-transitory computer-readable medium of claim 16 , further storing instructions which, when executed by the at least one processor, cause the at least one processor to generate the activity meta summary by:

generating a meta prompt comprising a prompt instruction to generate the activity meta summary using the synthesized activity items; and

providing the meta prompt and the synthesized activity items to the large language model.

19 . The non-transitory computer-readable medium of claim 16 , further storing instructions which, when executed by the at least one processor, cause the at least one processor to:

identify a second user account;

determine, within the activity data items, collaboration events between the user account and the second user account for the specified timeframe; and

identify a subset of synthesized activity items from the activity meta summary according to the collaboration events between the user account and the second user account for the specified timeframe.

20 . The non-transitory computer-readable medium of claim 16 , further storing instructions which, when executed by the at least one processor, cause the at least one processor to:

identify an operational initiative;

determine, within the activity data items, operational initiative events for the specified timeframe; and

identify a subset of synthesized activity items according to the operational initiative events for the specified timeframe.