Abstract: A method for generating an hallucination-free, retrospectively verified document on a specific subjects includes receiving an query related to the specific subject; generating a refined query from the query by applying a system query refiner and a user query refiner to the original query; generating an embedding of the refined query; applying the embedding to a vector database containing vectors representing data objects in the big data source; determining a similarity between the embedding and the vector representations. The method further includes applying the refined query and one or more most similar vectors to a large language model, applying the large language model to the big data source, and generating the verified document using data objects from the big data source, and returning the verified document and an identification for each of the data objects.