Semantic search and summarization for electronic documents

1 . A method, comprising:

receiving a selection signal from a graphical user interface (GUI) element of a GUI view, the selection signal to represent a request for a summary of an electronic document;

retrieving context information associated with the electronic document in response to the selection signal, the context information to comprise information representing a meaning or interpretation of document content within the electronic document;

generating a search query based on the context information, the search query expressed in a natural language representation to request a search for information within the electronic document;

generating a contextualized embedding for the search query to form a search vector;

retrieving a set of candidate document vectors that are semantically similar to the search vector from a document index of contextualized embeddings for the electronic document;

sending a request to a generative artificial intelligence (AI) model for an abstractive summary of document content for the set of candidate document vectors, the abstractive summary expressed in a natural language representation; and

receiving a response with the abstractive summary from the generative AI model.

2 . The method of claim 1 , wherein the context information comprises at least one of the following: a metadata for the electronic document, a location information for the electronic document, an access history for the electronic document, usage statistics for the electronic document, one or more hyperlinks to one or more other electronic documents, a content summary, one or more labels or tags for the electronic document, a semantic context for the electronic document, one or more document types associated with the electronic document, one or more identification data associated with the electronic document, one or more user(s) information associated with the electronic document, and any combinations thereof.

3 . The method of claim 1 , wherein the selection signal is generated upon actuation of the GUI element.

4 . The method of claim 3 , further comprising

receiving another selection signal from the GUI element or at least another GUI element of the GUI view;

in response to the receiving, forming one or more additional search queries based on at least one of: the context information associated with the electronic document, the abstractive summary of document content, and any combination thereof; and

presenting, on the GUI view, the one or more additional search queries.

5 . The method of claim 4 , further comprising

executing an additional search query in the one or more additional queries;

transmitting the additional search query to the one or more generative AI models for another abstractive summary of content in the electronic document responsive to the additional search query; and

presenting, on the GUI view, the another abstractive summary of the content received from the one or more generative AI models.

6 . The method of claim 1 , further comprising

receiving one or more additional search queries for information included within the electronic document, the one or more additional search queries being related to at least one of: the context information associated with the electronic document, the abstractive summary of document content, and any combination thereof; and

transmitting the one or more additional search queries to the one or more generative AI models for another abstractive summary of document content in the electronic document responsive to the one or more additional search queries; and

presenting, on the GUI view, the another abstractive summary of document content received from the one or more generative AI models.

7 . The method of claim 1 , wherein the abstractive summary of document content is generated based on at least one of the following: the context information associated with the electronic document, at least one external data source queried by the one or more generative AI models, and any combination thereof.

8 . The method of claim 1 , further comprising generating one or more table of contents associated with the electronic document based on at least one of the following: the context information associated with the electronic document, the abstractive summary of document content, and any combination thereof.

9 . The method of claim 1 , wherein the GUI view includes at least one of the following: one or more views of the electronic document, one or more views of text information associated with the electronic document, one or more views of graphical elements associated with one or more AI assistants, one or more views of one or more tables of contents associated with the electronic document, one or more views of common searches, one or more views of one or more defined search queries, one or more views of one or more document summaries associated with entireties of the electronic document, one or more views associated with one or more portions of the electronic document, one or more views of one or more feedback icons, one or more views of one or more search queries, one or more views of one or more text snippets associated with the electronic document and related to the abstractive summary of document content, and any combinations thereof.

10 . The method of claim 1 , wherein the contextualized embedding includes a vector representation of a sequence of words that includes contextual information for the sequence of words.

11 . The method of claim 1 , further comprising training a bidirectional encoder representation from transformers (BERT) language model composed of one or more transformer encoder layers using training data from the electronic document associated with one or more defined entities and having one or more electronic signatures.

12 . The method of claim 1 , further comprising generating the contextualized embeddings using one or more transformer architectures, the one or more transformer architectures including a bidirectional encoder representations from transformers (BERT) language model composed of one or more transformer encoder layers.

13 . The method of claim 1 , further comprising

generating the contextualized embeddings using a bidirectional encoder representations from transformers (BERT) language model;

indexing the contextualized embeddings for the electronic document to form the document index; and

storing the document index in at least one storage location.

14 . The method of claim 1 , wherein the contextualized embeddings include at least one of the following: one or more word level vectors, one or more sentence level vectors, one or more paragraph level vectors, and any combinations thereof.

15 . The method of claim 1 , further comprising retrieving the set of candidate document vectors that are semantically similar to the search vector using one or more semantic ranking algorithms.

16 . The method of claim 11 , wherein the electronic document includes at least one of the following: an unsigned electronic agreement with metadata including one or more signature tag marker element (STME) information suitable to receive one or more electronic signatures, a signed electronic document, and any combination thereof.

17 . A system, comprising:

at least one processor; and

at least one non-transitory storage media storing instructions, that when executed by the at least one processor, cause the at least one processor to

retrieve context information associated with an electronic document in response to a selection signal representing a request for a summary of an electronic document, the context information to comprise information representing a meaning or interpretation of document content within the electronic document, the selection signal received from a graphical user interface (GUI) element of a GUI view;

generate a search query based on the context information, the search query expressed in a natural language representation to request a search for information within the electronic document;

generate a contextualized embedding for the search query to form a search vector;

retrieve a set of candidate document vectors that are semantically similar to the search vector from a document index of contextualized embeddings for the electronic document;

send a request to a generative artificial intelligence (AI) model for an abstractive summary of document content for the set of candidate document vectors, the abstractive summary expressed in a natural language representation; and

receive a response with the abstractive summary from the generative AI model.

18 . The system of claim 17 , wherein the context information comprises at least one of the following: a metadata for the electronic document, a location information for the electronic document, an access history for the electronic document, usage statistics for the electronic document, one or more hyperlinks to one or more other electronic documents, a content summary, one or more labels or tags for the electronic document, a semantic context for the electronic document, one or more document types associated with the electronic document, one or more identification data associated with the electronic document, one or more user(s) information associated with the electronic document, and any combinations thereof.

19 . The system of claim 17 , wherein the at least one processor is configured to

receive another selection signal from the GUI element or at least another GUI element of the GUI view;

in response to receiving the another selection, form one or more additional search queries based on at least one of: the context information associated with the electronic document, the abstractive summary of document content, and any combination thereof;

execute an additional search query in the one or more additional queries;

transmit the additional search query to the one or more generative AI models for another abstractive summary of content in the electronic document responsive to the additional search query; and

present, on the GUI view, the another abstractive summary of the content received from the one or more generative AI models.

20 . A computer program product comprising a non-transitory machine-readable medium storing instructions that, when executed by at least one programmable processor, cause the at least one programmable processor to:

generate a search query based on a context information, the search query expressed in a natural language representation to request a search for information within an electronic document, the context information being associated with an electronic document and includes information representing a meaning or interpretation of document content within the electronic document;

generate a contextualized embedding for the search query to form a search vector;

retrieve a set of candidate document vectors that are semantically similar to the search vector from a document index of contextualized embeddings for the electronic document;

send a request to a generative artificial intelligence (AI) model for an abstractive summary of document content for the set of candidate document vectors, the abstractive summary expressed in a natural language representation; and

receive a response with the abstractive summary from the generative AI model.