Incremental Summarization for Customer Support via Progressive Note-Taking and Agent Feedback

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Wu, Yisha, Zhao, Cen Mia, Cao, Yuanpei, Su, Xiaoqing, Mehdad, Yashar, Ji, Mindy, Cheng, Claire Na
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866908583267926016
author Wu, Yisha
Zhao, Cen Mia
Cao, Yuanpei
Su, Xiaoqing
Mehdad, Yashar
Ji, Mindy
Cheng, Claire Na
author_facet Wu, Yisha
Zhao, Cen Mia
Cao, Yuanpei
Su, Xiaoqing
Mehdad, Yashar
Ji, Mindy
Cheng, Claire Na
contents We introduce an incremental summarization system for customer support agents that intelligently determines when to generate concise bullet notes during conversations, reducing agents' context-switching effort and redundant review. Our approach combines a fine-tuned Mixtral-8x7B model for continuous note generation with a DeBERTa-based classifier to filter trivial content. Agent edits refine the online notes generation and regularly inform offline model retraining, closing the agent edits feedback loop. Deployed in production, our system achieved a 3% reduction in case handling time compared to bulk summarization (with reductions of up to 9% in highly complex cases), alongside high agent satisfaction ratings from surveys. These results demonstrate that incremental summarization with continuous feedback effectively enhances summary quality and agent productivity at scale.
format Preprint
id arxiv_https___arxiv_org_abs_2510_06677
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Incremental Summarization for Customer Support via Progressive Note-Taking and Agent Feedback
Wu, Yisha
Zhao, Cen Mia
Cao, Yuanpei
Su, Xiaoqing
Mehdad, Yashar
Ji, Mindy
Cheng, Claire Na
Computation and Language
Artificial Intelligence
Machine Learning
We introduce an incremental summarization system for customer support agents that intelligently determines when to generate concise bullet notes during conversations, reducing agents' context-switching effort and redundant review. Our approach combines a fine-tuned Mixtral-8x7B model for continuous note generation with a DeBERTa-based classifier to filter trivial content. Agent edits refine the online notes generation and regularly inform offline model retraining, closing the agent edits feedback loop. Deployed in production, our system achieved a 3% reduction in case handling time compared to bulk summarization (with reductions of up to 9% in highly complex cases), alongside high agent satisfaction ratings from surveys. These results demonstrate that incremental summarization with continuous feedback effectively enhances summary quality and agent productivity at scale.
title Incremental Summarization for Customer Support via Progressive Note-Taking and Agent Feedback
topic Computation and Language
Artificial Intelligence
Machine Learning
url https://arxiv.org/abs/2510.06677