Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Zhao, Cen Mia, Zhang, Tiantian, Su, Hanchen, Zhang, Yufeng Wayne, Su, Shaowei, Xu, Mingzhi, Liu, Yu Elaine, Han, Wei, Werner, Jeremy, Cheng, Claire Na, Mehdad, Yashar
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2510.06674
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866915540753186816
author Zhao, Cen Mia
Zhang, Tiantian
Su, Hanchen
Zhang, Yufeng Wayne
Su, Shaowei
Xu, Mingzhi
Liu, Yu Elaine
Han, Wei
Werner, Jeremy
Cheng, Claire Na
Mehdad, Yashar
author_facet Zhao, Cen Mia
Zhang, Tiantian
Su, Hanchen
Zhang, Yufeng Wayne
Su, Shaowei
Xu, Mingzhi
Liu, Yu Elaine
Han, Wei
Werner, Jeremy
Cheng, Claire Na
Mehdad, Yashar
contents We introduce an Agent-in-the-Loop (AITL) framework that implements a continuous data flywheel for iteratively improving an LLM-based customer support system. Unlike standard offline approaches that rely on batch annotations, AITL integrates four key types of annotations directly into live customer operations: (1) pairwise response preferences, (2) agent adoption and rationales, (3) knowledge relevance checks, and (4) identification of missing knowledge. These feedback signals seamlessly feed back into models' updates, reducing retraining cycles from months to weeks. Our production pilot involving US-based customer support agents demonstrated significant improvements in retrieval accuracy (+11.7% recall@75, +14.8% precision@8), generation quality (+8.4% helpfulness) and agent adoption rates (+4.5%). These results underscore the effectiveness of embedding human feedback loops directly into operational workflows to continuously refine LLM-based customer support system.
format Preprint
id arxiv_https___arxiv_org_abs_2510_06674
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Agent-in-the-Loop: A Data Flywheel for Continuous Improvement in LLM-based Customer Support
Zhao, Cen Mia
Zhang, Tiantian
Su, Hanchen
Zhang, Yufeng Wayne
Su, Shaowei
Xu, Mingzhi
Liu, Yu Elaine
Han, Wei
Werner, Jeremy
Cheng, Claire Na
Mehdad, Yashar
Artificial Intelligence
We introduce an Agent-in-the-Loop (AITL) framework that implements a continuous data flywheel for iteratively improving an LLM-based customer support system. Unlike standard offline approaches that rely on batch annotations, AITL integrates four key types of annotations directly into live customer operations: (1) pairwise response preferences, (2) agent adoption and rationales, (3) knowledge relevance checks, and (4) identification of missing knowledge. These feedback signals seamlessly feed back into models' updates, reducing retraining cycles from months to weeks. Our production pilot involving US-based customer support agents demonstrated significant improvements in retrieval accuracy (+11.7% recall@75, +14.8% precision@8), generation quality (+8.4% helpfulness) and agent adoption rates (+4.5%). These results underscore the effectiveness of embedding human feedback loops directly into operational workflows to continuously refine LLM-based customer support system.
title Agent-in-the-Loop: A Data Flywheel for Continuous Improvement in LLM-based Customer Support
topic Artificial Intelligence
url https://arxiv.org/abs/2510.06674