By Chutian Wang, Zhiheng Xu, Paul Lou, Ziyi Wang, Jiayu Lou, Liuming Zhang, Jingwen Qiang, Clint Kelly, Lei Shi, Dan Zhao, Xu Hu, Jianqi Liao, Zecheng Xu, Tong Chen Artificial intelligence and large…

Airbnb's platform is  dedicated to providing insights and resources for developers and technology enthusiasts, focusing on engineering challenges, innovation projects, and best practices in software development. Through articles, tech talks, and open-source contributions, Airbnb offers insights into building scalable and reliable technology solutions. Developers can learn about Airbnb's engineering culture, technology stack, and contributions to the broader tech community.

Airbnb

Airbnb has upgraded its Automation Platform from version 1 to version 2 to improve conversational AI using large language models (LLMs). The new version facilitates LLM applications that enhance customer support by allowing more natural and efficient interactions. Key features include context management, guardrail frameworks to ensure ethical communication, and the Chain of Thought workflow, which improves AI’s reasoning and problem-solving abilities. Although LLMs offer significant advancements, Airbnb balances their use with traditional workflows to address current limitations such as latency and strict validation needs.

Automation Platform v2: Improving Conversational AI at Airbnb

Challenges of Traditional Conversational AI Systems

Opportunities of Conversational AI Driven by LLM

Architecture of LLM Application on Automation Platform v2