BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//pretalx//global2024.pydata.org//Y7FKDD
BEGIN:VEVENT
UID:pretalx-cfp-Y7FKDD@global2024.pydata.org
DTSTART:20241204T130000Z
DTEND:20241204T133000Z
DESCRIPTION:Traditional document processing for Retrieval-Augmented Generat
 ion (RAG) often involves cumbersome\, error-prone extraction pipelines\, h
 ampering AI's ability to retrieve high-quality information from complex fo
 rmats like PDFs and PowerPoint decks. ColPali disrupts this process by emb
 edding entire pages—text\, visuals\, and layout—into rich\, multi-vect
 or representations using Vision Language Models (VLMs). This talk explores
  how ColPali\, paired with multimodal models like the Llama 3.2 Vision ser
 ies\, enables RAG systems to “see” and reason over documents\, dramati
 cally improving retrieval performance. Attendees will learn to implement C
 olPali for enhanced\, scalable\, and robust enterprise knowledge retrieval
 .
DTSTAMP:20250709T220237Z
LOCATION:General Track
SUMMARY:Breaking Free from Extraction Pipelines: ColPali’s Vision-Powered
  RAG for Enterprise Documents - Zain Hasan
URL:https://global2024.pydata.org/cfp/talk/Y7FKDD/
END:VEVENT
END:VCALENDAR