Proposagen: AI-Driven Automated Proposal Generation System

The current process of generating proposals for clients is time-consuming and generic, with no personalization. This paper introduces ProposaGen, an AI-driven system that automates and personalises the proposal generation process using Large Language Models (LLMs) and the Trans-former architecture. It uses fine-tuned LLaMA models as a base, LangChain for workflow automation, and DEMON-STRATE-SEARCH-PREDICT (DSPy) for optimization. It also features a chatbot for interaction with the client. Our findings indicate that ProposaGen significantly accelerates proposal generation process while reaching a performance comparable to manual methods in key areas such as consistency and covering requirements. Although certain metrics show room for improvement due to the technical complexity of engineering projects, ProposaGen is able to ease workloads, making it a valuable tool for businesses seeking to enhance efficiency.