Systems and methods for querying a document pool

A system comprising at least one computer including a processor, wherein the at least one computer is configured to:

• receive, at a query platform, a query text string;
• tokenize, by the query platform, the query text string, wherein the tokenizing generates a query vector embedding;
• determine, by the query platform, a document vector embedding by numerically representing a semantic meaning of text of a group of tokens from a document;
• compare the document vector embedding to the query vector embedding and determining the comparison is above a similarity threshold;
• retrieve, by the query platform, textual document data of the document related to the document vector embedding by using a document identifier from a datastore, wherein the document identifier interrelates the document, a token of the group of tokens from the document, and the document vector embedding;
• determine, by a similarity engine of the query platform, a similar document vector embedding based on the document vector embedding;
• send, by the query platform, an input comprising the query text string, the textual document data, the document embedding vector, and the similar document vector embedding to a generative model engine;
• expose, by the generative model engine, an instructional parameter that instructs the generative model engine to generate a response using only documents associated with the group of tokens;
• receive, by the query platform and from the generative model engine, a natural language response to the input;
• and
• display, by the query platform, the natural language response via an interface.

A non-transitory computer readable storage medium, including instructions stored thereon, which instructions, when read and executed by one or more computer processors, cause the one or more computer processors to perform steps comprising:

• receiving, at a query platform, a query text string;
• tokenizing, by the query platform, the query text string, wherein the tokenizing generates a query vector embedding;
• determining, by the query platform, a document vector embedding by numerically representing a semantic meaning of text of a group of tokens from a document;
• comparing the document vector embedding to the query vector embedding and determining the comparison is above a similarity threshold;
• retrieving, by the query platform, textual document data of the document related to the document vector embedding by using a document identifier from a datastore, wherein the document identifier interrelates the document, a token of the group of tokens from the document, and the document vector embedding;
• determining, by a similarity engine of the query platform, a similar document vector embedding based on the document vector embedding;
• sending, by the query platform, an input comprising the query text string, the textual document data, the document embedding vector, and the similar document vector embedding to a generative model engine;
• exposing, by the generative model engine, an instructional parameter that instructs the generative model engine to generate a response using only documents associated with the document and one or more documents associated with the group of tokens;
• receiving, by the query platform and from the generative model engine, a natural language response to the input;
• and
• displaying, by the query platform, the natural language response via an interface.