Скачать или смотреть DSS2023 Nick Flores GE "Large-Language Models – Challenges and Opportunities in Business Use"

While the popularization of ChatGPT and the rapid evolution of Generative AI tools promise to revolutionize many industries, businesses face many practical challenges in implementing this new technology across the enterprise. In this talk we will discuss some of the key issues and best practices related to LLMs, including IP, privacy, and ethical concerns, building trust & quality in the models, conducting fine-tuning and evaluation, integrating with data sources, and measuring/optimizing ROI. The GE Aerospace team will share several real-world examples and case studies that highlight the challenges we have overcome from business, functional, and infrastructure perspectives.

Bio: Nick Flores is a Sr. Staff Data Scientist at GE Aerospace. Over his 15 year career with GE, including 11 years with GE Aerospace, he transitioned from the field of mechanical design engineering to data science. As a data scientist at GE Aerospace, he has leveraged his deep domain expertise to unlock data-driven insights for supply chain machine health and engine forecasting initiatives. Recently, he has led an initiative to develop a customized Large Language Model (LLM) with internal company documents.