Method and system for optimizing generative artificial intelligence-based multimodel conversational environment

Abstract

Disclosed herein is a method for Generative Artificial Intelligence (GenAI)-based one or more multi-model conversational environments. The method comprises obtaining, an input message from a conversational environment, selecting, a language model for the input message, generating, via the selected language model, a response message for the input message, obtaining at least one piece of user feedback of the response message from a message stream of the conversational environment in real-time, determining an efficacy of the selected language model based on the user feedback, estimating, by retrieving one or more parameters from a database, and optimizing the efficacy of the selected language model by adjusting the predetermined weights of the selected language model based on the value optimization parameter.

Images