Скачать с ютуб LlamaIndex Webinar: Document Metadata and Local Models for Better, Faster Retrieval в хорошем качестве

LlamaIndex Webinar: Document Metadata and Local Models for Better, Faster Retrieval

​In this webinar we're excited to bring on members from Docugami to talk about some crucial concepts for better RAG systems: - ​Using document metadata for high accuracy retrieval. Document metadata has proven crucial in injecting context. - ​Using local (not OpenAI etc) models for retrieval, chat, and agents. This may or may not improve security (there's a public perception that OSS models are more secure, whether or not that's completely true), but can also improve latency/cost-control. ​Presenters: - ​Taqi Jaffri (co-founder, Head of Product @ Docugami) - ​Mike Palmer (co-founder, Head of Technologies @ Docugami) - ​Zubin Wadia (PM @ Docugami)