Building productive AIs with your documents - Done right

Sooner or later, businesses will enable employees to interact with internal documentation through an AI overlay (Retrieval Augmented Generation). While it sounds (deceptively) as simple as "upload documents as embedding behind an OpenAI key," it's not. As shown by multiple interactions we've had with startups and mature businesses who have already worked on this, pulling code from GitHub and plugging an LLM API key is not the part that truly matters, And it's the only part that's fun. Gabriel will share the lessons from building a unified retrieval engine scaled on millions of documents, helping through multiple use cases and how to leverage existing technologies best to create a PoC while being cognizant of downstream technological invisibilities. Participants will gain insightful clues to help them better map out any RAG projects.

La conférence est présentée par Gabriel Lespérance.