SYSTEMS AND METHODS FOR ZERO-SHOT, FAST-GENERATION AND IMPLEMENTATION OF AN INTELLIGENT VIRTUAL DIALOGUE AGENT USING ONE OR MORE PRE-TRAINED MACHINE LEARNING-BASED LANGUAGE MODELS AND A RESPONSE CORPUS

System and method of building a zero-shot training, machine learning-based virtual dialogue agent includes identifying a response corpus comprising a plurality of distinct response samples; providing the response corpus to a virtual agent generator; generating, by one or more pre-trained machine learning language models, a distinct embeddings inference for each of the plurality of distinct response samples; forming an embeddings-based architecture for response generation based on the distinct embeddings inference for each of the plurality of distinct response samples, wherein the embeddings-based architecture includes a mapping of the distinct embeddings inference for each of the plurality of distinct response samples to an n-dimensional space; instantiating a virtual dialogue agent based on receiving user stimuli; and computing, using the embeddings-based architecture, a response inference to the user stimuli, wherein the computed response inference is based on identifying one distinct embeddings inference of a distinct response sample of the response corpus.