Advances and applications of large language models II

Abstract

In today's fast-paced business environment, efficient meeting documentation is crucial. However, manual transcription and minute-taking can be time-consuming and error-prone. This project developed a python web application – Convo2Calendar , that can convert meeting audio recordings into formatted meeting notes, provide meeting analytics and allow users to chat with their documents. The Streamlit demo walkthrough is available at https://youtu.be/rrsmBsllBaE.

This project leverages advanced Large Language Models (LLMs), renowned for their capabilities in complex natural language processing (NLP) tasks. By integrating LLMs with other libraries, Convo2Calendar functions as a comprehensive 5-in-1 virtual meeting minute writer, including a bias-free transcription and a user-friendly 24/7 Q&A service.