In this talk, we will discuss the lessons learned from using Large Language Models (LLMs) at Hyro, a company that specializes in conversational AI.

We will introduce Hyro’s approach and explain how our “hybrid” architecture helped us incorporate LLMs, into Healthcare and beyond. We will dive into the considerations using LLMs in production, such as price, privacy, and hallucinations.

We will also talk about productization of RAG (retrieval augmented generation) solutions – retrieval methods, prompt engineering, evaluation and control. We will learn how to build a stable, reliable, and good quality system to enhance NLU capabilities across the board.