document-ai.

NeoMME: Why Unified Multimodal Encoders Matter for Document AI

NeoMME ditches separate vision and text encoders for unified multimodal retrieval. What this means for building faster, leaner document search systems.