Record snapshot

1 . A computer-implemented method, comprising:

parsing a data record responsive to a selection of the data record by a user, by at least one computer processor, into a plurality of data fields;

generating, by the at least one computer processor, a hierarchical tree graph of context data based on the plurality of data fields and corresponding to a prompt template selected based on a type of a prompt request, wherein the data record is identified based on a data-record reference in the prompt request;

generating a prompt from the prompt template by grounding the prompt template with a data object comprising the hierarchical tree graph of context data, the grounding including entering data from the data record into corresponding fields of the prompt template; and

querying a large language model with the prompt to generate an output specific to the data record.

2 . The method of claim 1 , wherein parsing the data record to generate the hierarchical tree graph of context data comprises:

extracting a plurality of fields of the data record based on a page layout of the data record; and

generating, based on the extracted plurality of fields, the hierarchical tree graph of context data corresponding to the page layout of the data record.

3 . The method of claim 1 , wherein identifying the data record comprises:

matching a characteristic of the data record to the data-record reference in the prompt request.

4 . The method of claim 1 , wherein generating the prompt from the prompt template comprises:

resolving a plurality of merge fields of the prompt template using the data object.

5 . The method of claim 1 , wherein the prompt template is grounded by embedding the data object within the prompt template in a single merge field.

6 . The method of claim 1 , wherein the data record is customer relationship management data.

7 . The method of claim 1 , wherein the type of the prompt request is a request to generate a type of email based on the data record.

8 . The method of claim 1 , wherein the data object is a JavaScript object notation (JSON) object.

9 . A system, comprising:

a memory; and

at least one processor coupled to the memory and configured to perform operations comprising:

parsing a data record responsive to a selection of the data record by a user into a plurality of data fields;

generating a hierarchical tree graph of context data based on the plurality of data fields and corresponding to a prompt template selected based on a type of a prompt request, wherein the data record is identified based on a data-record reference in the prompt request,

generating a prompt from the prompt template by grounding the prompt template with a data object comprising the hierarchical tree graph of context data, the grounding including entering data from the data record into corresponding fields of the prompt template; and

querying a large language model with the prompt to generate an output specific to the data record.

10 . The system of claim 9 , wherein to parse the data record to generate the hierarchical tree graph of context data, the operations further comprise:

extracting a plurality of fields of the data record based on a page layout of the data record; and

generating, based on the extracted plurality of fields, the hierarchical tree graph of context data corresponding to the page layout of the data record.

11 . The system of claim 9 , to the operations further comprising:

matching a characteristic of the data record to the data-record reference in the prompt request.

12 . The system of claim 9 , wherein to generate the prompt from the prompt template, the operations further comprising:

resolving a plurality of merge fields of the prompt template using the data object.

13 . The system of claim 9 , wherein the prompt template is grounded by embedding the data object within the prompt template in a single merge field.

14 . The system of claim 9 , wherein the data record is customer relationship management data.

15 . A non-transitory computer-readable medium (CRM) having instructions stored thereon that, when executed by at least one computing device, causes the at least one computing device to perform operations comprising:

parsing a data record responsive to a selection of the data record by a user into a plurality of data fields;

generating a hierarchical tree graph of context data based on the plurality of data fields and corresponding to a prompt template selected based on a type of a prompt request, wherein the data record is identified based on a data-record reference in the prompt request;

generating a prompt from the prompt template by grounding the prompt template with a data object comprising the hierarchical tree graph of context data, the grounding including entering data from the data record into corresponding fields of the prompt template; and

querying a large language model with the prompt to generate an output specific to the data record.

16 . The non-transitory CRM of claim 15 , wherein parsing the data record to generate the hierarchical tree graph of context data comprises:

extracting a plurality of fields of the data record based on a page layout of the data record; and

generating, based on the extracted plurality of fields, the hierarchical tree graph of context data corresponding to the page layout of the data record.

17 . The non-transitory CRM of claim 15 , wherein identifying the data record comprises:

matching a characteristic of the data record to the data record reference in the prompt request.

18 . The non-transitory CRM of claim 15 , wherein generating the prompt from the prompt template comprises:

resolving a plurality of merge fields of the prompt template using the data object.

19 . The non-transitory CRM of claim 15 , wherein the prompt template is grounded by embedding the data object within the prompt template in a single merge field.

20 . The non-transitory CRM of claim 15 , wherein the data record is customer relationship management data.