Building a Multimodal LLM-Based Search Assistant Chatbot for Housing Search - Haystack EU 2024

Explore the development of a cutting-edge multimodal LLM-based search assistant chatbot designed to revolutionize the housing search process. Dive into QuintoAndar Group's innovative approach, which combines contrastive vision-language models and Large Language Models (LLMs) to create a conversational interface that enhances user experience. Learn how this Latin American housing platform leader leverages AI technologies to streamline searches, reduce paperwork, and increase accessibility. Discover the adopted architecture, engineering practices for working with LLMs, and the integration of traditional search mechanisms with multimodal capabilities. Gain insights from industry experts Tetiana Torovets, Giulio Cesare Mastrocinque Santo, and Lucas Cardozo as they share their experiences in developing this groundbreaking search assistant, discussing challenges faced and solutions implemented in this 41-minute presentation from Haystack EU 2024.