Enterprise group chat memory structure improves information retrieval accuracy

EGCForest: Structured Agent Memory Organization for Enterprise Group Chats

Computation and Language

Summary

Enterprise chats have lots of important information but it's hard to find because messages are spread out over many channels and come from different people. The authors created EGCForest, which organizes chat messages into three connected trees based on channel, speaker, and project deliverables. This helps find facts by looking at how messages relate in time, who said them, and how they connect across the chat history. Their tests show this method is more accurate and efficient than previous ways to search these chats.

What this means in practice

  • For enterprise software engineers: Build improved chat search tools that find project details efficiently by using multi-view memory structures to organize and retrieve messages.
  • For customer support teams: Better track and retrieve conversation history related to customer issues spread across multiple chat channels and agents for faster resolution.

Authors

Lingxi Cui, Jinhao Ruan, Wenqing Wei, Guo Yi, Li Rao, Ke Chen, Lidan Shou

Abstract

Enterprise group chats record project progress, decisions, task ownership, and other organizational information valuable to enterprise agents. Reliable retrieval is challenging because evidence spans interleaved threads, channels, and time, with meaning and authority conditioned on the speaker. Neither RAG's independent chunk retrieval nor memory systems' summary-based consolidation can reliably recover such distributed evidence. We present EGCForest, a novel multi-view hierarchical memory framework for enterprise group chats. EGCForest organizes an immutable message log into three complementary forests: a Channel forest for temporal workflows, an Actor forest for speaker-conditioned context, and an Artifact forest for deliverable lifecycles. Cross-view connections are enabled by a Placement Map and sparse structural bridges. Built on this memory backbone, EGCForest retrieves and assembles evidence by jointly leveraging message-level relevance, hierarchical context from forest summaries, and cross-view links, recovering distributed evidence. Experiments on three enterprise-chat benchmarks show that EGCForest consistently outperforms RAG and memory-system baselines, improving answer accuracy by up to 20.85 points over MemOS while supporting efficient online retrieval.