@inproceedings{zhao-etal-2025-agent,
    title = "Agent-in-the-Loop: A Data Flywheel for Continuous Improvement in {LLM}-based Customer Support",
    author = "Zhao, Cen  and
      Zhang, Tiantian  and
      Su, Hanchen  and
      Zhang, Yufeng  and
      Su, Shaowei  and
      Xu, Mingzhi  and
      Liu, Yu  and
      Han, Wei  and
      Werner, Jeremy  and
      Cheng, Claire Na  and
      Mehdad, Yashar",
    editor = "Potdar, Saloni  and
      Rojas-Barahona, Lina  and
      Montella, Sebastien",
    booktitle = "Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing: Industry Track",
    month = nov,
    year = "2025",
    address = "Suzhou (China)",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/ingest-emnlp/2025.emnlp-industry.135/",
    pages = "1919--1930",
    ISBN = "979-8-89176-333-3",
    abstract = "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."
}