Description:

Unlock the full potential of AI by integrating multiple Large Language Models (LLMs) into a cohesive and powerful Multi-LLM Copilot. In this session, we will dive deep into a real-world case study, showcasing how different LLMs can be orchestrated to perform distinct tasks, creating a sophisticated and versatile assistant.

We'll explore the complete design process, highlighting key components with below outline

Introduction (2 mins) - Introduce self and Copilot case Why Multi-LLM Copilots? (5 mins) - Why we need multi-LLMs in the copilot and its advantages. Multi-LLM Copilot in Action (18 mins) - 1. Detailed component analysis - NL2SQL, Intent Detection, Summarization and contextual augmentation. 2. How multi-LLM setup helped improving the performance and quality. Questions and Answers (5 mins) - Discuss questions.

Prerequisites:

Basic understanding of LLMs, RAG, Prompt Engineering and Python.

Speaker Info:

With over 13 years of expertise in Generative AI, Machine Learning, NLP, Deep Learning, and Data Analytics, I currently serve as a Senior Data and Applied Scientist at Microsoft. My work focuses on developing Copilots and advanced ML solutions to empower Microsoft Sellers, Partners, and customers.

Passionate about sharing knowledge, I’ve been actively teaching, mentoring, and engaging with the community for the past five years through various platforms. I hold a master’s degree in Data Science and Engineering from Birla Institute of Technology and Science, Pilani.

Speaker Links:

Linkedin Profile- https://www.linkedin.com/in/agnitin/ Recent Session on Prompt Engineering - https://www.mygreatlearning.com/webinars/recording?token=OTI0OTk2MzE5Njg=&prospectId=4a41ad34-d6b9-4d99-91b9-a435d6d300eb Blog - https://cognibits.hashnode.dev/