Computer-generated content based on text classification, semantic relevance, and activation of deep learning large language models

The disclosure relates to automatically generating unique content including natural language text based on a corpus of previously generated response documents and discrete requirements defined in a requirements specification. Embodiments use a large language model (LLM) 150 to be able to generate a proposal to meet specific requirements (e.g. for a contractor in response to a procurement call with specific requirements). For each requirement in the requirements text relevant sections in a store of response sections 123 are identified and a language model matrix is generated. The language model matrix comprises natural language text from the identified responses and the requirement to generate unique natural language text. This language model matrix is used with a pre-trained deep learning language model to generate candidate response sections. A final response document is compiled from candidate responses for each requirement. The system may use generative stitching 160 that includes multi-layer processes that execute to influence the generation of unique content including natural language text through an artificial intelligence (AI) language transformer model trained to output the content based on previously written material that is semantically relevant to the discrete requirements and is weighted against labelled attributes.