Generative artificial intelligence platform to manage smart documents

1 . A method, comprising:

retrieving an unsigned electronic agreement associated with a defined entity, the unsigned electronic agreement having a set of document clauses, each document clause to have one or more parameters;

retrieving a candidate source associated with the defined entity, the candidate source having a set of candidate clauses, each candidate clause to have one or more parameters;

identifying a set of related document clauses between the unsigned electronic agreement and the candidate source using a machine learning model, the set of related document clauses to comprise a document clause from the set of documents clauses of the unsigned electronic agreement and a candidate clause from the set of candidate clauses of the candidate source;

identifying a set of common parameters shared between the set of related document clauses using the machine learning model, the set of common parameters to include a parameter from the document clause and a parameter from the candidate clause of the set of related document clauses; and

generating an insight based on a formal deviation between the set of common parameters using a natural language generation (NLG) model, the insight including a text summary generated by a summarization component of the NLG model and describing the formal deviation in a natural language representation.

2 . The method of claim 1 , further comprising presenting a description of the formal deviation as natural language text on an electronic display.

3 . The method of claim 1 , further comprising reproducing a description of the formal deviation as synthesized speech over an electronic speaker.

4 . The method of claim 1 , wherein the candidate source to comprise a document rule set associated with the defined entity, the document rule set to include a set of document rules for the unsigned electronic agreement, each document rule associated with a candidate clause.

5 . The method of claim 1 , wherein the unsigned electronic agreement is a first document version of the unsigned electronic agreement and the candidate source is a second document version of the unsigned electronic agreement, the second document version of the unsigned electronic agreement to share a structure and one or more candidate clauses with the first document version of the unsigned electronic agreement at a defined point in time, wherein a candidate clause of the second document version is modified relative to the first document version.

6 . The method of claim 1 , wherein the candidate source to comprise a signed electronic agreement, the signed electronic agreement associated with the defined entity and having an electronic signature.

7 . The method of claim 1 , further comprising:

generating a similarity score for each candidate clause of the set of candidate clauses of the candidate source using the machine learning model;

generating a similarity score for each document clause of the set of documents clauses of the unsigned electronic agreement using the machine learning model; and

identifying the set of related document clauses shared between the set of documents clauses of the unsigned electronic agreement and the set of candidate clauses of the candidate source based on the similarity scores.

8 . A computing apparatus, comprising:

a processing circuitry; and

a memory communicatively coupled to the processing circuitry, the memory storing instructions that, when executed by the processing circuitry, cause the processing circuitry to:

retrieve an unsigned electronic agreement associated with a defined entity, the unsigned electronic agreement having a set of document clauses, each document clause to have one or more parameters;

retrieve a candidate source associated with the defined entity, the candidate source having a set of candidate clauses, each candidate clause to have one or more parameters;

identify a set of related document clauses between the unsigned electronic agreement and the candidate source using a machine learning model, the set of related document clauses to comprise a document clause from the set of documents clauses of the unsigned electronic agreement and a candidate clause from the set of candidate clauses of the candidate source;

identify a set of common parameters shared between the set of related document clauses using the machine learning model, the set of common parameters to include a parameter from the document clause and a parameter from the candidate clause of the set of related document clauses; and

generate an insight based on a formal deviation between the set of common parameters using a natural language generation (NLG) model, the insight including a text summary generated by a summarization component of the NLG model and describing the formal deviation in a natural language representation.

9 . The computing apparatus of claim 8 , wherein the processing circuitry is configured to present a description of the formal deviation as natural language text on an electronic display.

10 . The computing apparatus of claim 8 , wherein the processing circuitry is configured to reproduce a description of the formal deviation as synthesized speech over an electronic speaker.

11 . The computing apparatus of claim 8 , wherein the candidate source to comprise a document rule set associated with the defined entity, the document rule set to include a set of document rules for the unsigned electronic agreement, each document rule associated with a candidate clause.

12 . The computing apparatus of claim 8 , wherein the unsigned electronic agreement is a first document version of the unsigned electronic agreement and the candidate source is a second document version of the unsigned electronic agreement, the second document version of the unsigned electronic agreement to share a structure and one or more candidate clauses with the first document version of the unsigned electronic agreement at a defined point in time, wherein a candidate clause of the second document version is modified relative to the first document version.

13 . The computing apparatus of claim 8 , wherein the candidate source to comprise a signed electronic agreement, the signed electronic agreement associated with the defined entity and having an electronic signature.

14 . The computing apparatus of claim 8 , wherein the processing circuitry is configured to:

generate a similarity score for each candidate clause of the set of candidate clauses of the candidate source using the machine learning model;

generate a similarity score for each document clause of the set of documents clauses of the unsigned electronic agreement using the machine learning model; and

identify the set of related document clauses shared between the set of documents clauses of the unsigned electronic agreement and the set of candidate clauses of the candidate source based on the similarity scores.

15 . A non-transitory computer-readable storage medium, the computer-readable storage medium including instructions that when executed by a computer, cause the computer to:

retrieve an unsigned electronic agreement associated with a defined entity, the unsigned electronic agreement having a set of document clauses, each document clause to have one or more parameters;

retrieve a candidate source associated with the defined entity, the candidate source having a set of candidate clauses, each candidate clause to have one or more parameters;

identify a set of related document clauses between the unsigned electronic agreement and the candidate source using a machine learning model, the set of related document clauses to comprise a document clause from the set of documents clauses of the unsigned electronic agreement and a candidate clause from the set of candidate clauses of the candidate source;

identify a set of common parameters shared between the set of related document clauses using the machine learning model, the set of common parameters to include a parameter from the document clause and a parameter from the candidate clause of the set of related document clauses; and

generate an insight based on a formal deviation between the set of common parameters using a natural language generation (NLG) model, the insight including a text summary generated by a summarization component of the NLG model and describing the formal deviation in a natural language representation.

16 . The computer-readable storage medium of claim 15 , the computer is configured to present a description of the formal deviation as natural language text on an electronic display.

17 . The computer-readable storage medium of claim 15 , the computer is configured to reproduce a description of the formal deviation as synthesized speech over an electronic speaker.

18 . The computer-readable storage medium of claim 15 , wherein the candidate source to comprise a document rule set associated with the defined entity, the document rule set to include a set of document rules for the unsigned electronic agreement, each document rule associated with a candidate clause.

19 . The computer-readable storage medium of claim 15 , wherein the unsigned electronic agreement is a first document version of the unsigned electronic agreement and the candidate source is a second document version of the unsigned electronic agreement, the second document version of the unsigned electronic agreement to share a structure and one or more candidate clauses with the first document version of the unsigned electronic agreement at a defined point in time, wherein a candidate clause of the second document version is modified relative to the first document version.

20 . The computer-readable storage medium of claim 15 , wherein the candidate source to comprise a signed electronic agreement, the signed electronic agreement associated with the defined entity and having an electronic signature.