Definition
The process by which an AI agent or RAG system synthesizes, compares, and aggregates information distributed across multiple retrieved documents to answer complex, multi-hop queries. While it enables deep analytical insights, it introduces trade-offs such as increased latency due to multi-step reasoning and higher token costs from larger context requirements.
Not simple multi-document retrieval, but the active logical synthesis of those sources.
"A detective's corkboard with red strings connecting facts across different case files."
Conceptual Overview
The process by which an AI agent or RAG system synthesizes, compares, and aggregates information distributed across multiple retrieved documents to answer complex, multi-hop queries. While it enables deep analytical insights, it introduces trade-offs such as increased latency due to multi-step reasoning and higher token costs from larger context requirements.
Disambiguation
Not simple multi-document retrieval, but the active logical synthesis of those sources.
Visual Analog
A detective's corkboard with red strings connecting facts across different case files.