@inproceedings{wang-etal-2026-ragpulse,
title = "{RAGP}ulse: A Real-World {RAG} Workload Capturing Authentic Temporal Reuse Patterns for Efficient Serving",
author = "Wang, Zhengchao and
Hu, Yitao and
Chang, Zhuxuan and
Ye, Jianing and
Yang, Guotao and
Lai, Tianhao and
Yu, Jiazheng and
Zhao, Zhixin and
Zheng, Liang and
Youpeng, Deng and
Chen, Sheng and
Li, Keqiu",
editor = "Lee, Hung-Yi and
Dev, Sunipa and
Li, Jessy and
Elliott, Desmond",
booktitle = "Findings of the {A}ssociation for {C}omputational {L}inguistics: {EMNLP} 2026",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.findings-emnlp.335/",
pages = "7065--7076",
ISBN = "979-8-89176-462-0",
abstract = "Retrieval-augmented generation (RAG) has emerged as a critical paradigm for deploying reliable and knowledge-intensive large language model (LLM) applications. Its unique execution pipeline introduces large overhead in system retrieval, such as massive context loading and document dependencies. To reduce this overhead, existing work on RAG serving systems often relies on generic LLM inference traces or synthetic RAG datasets. However, these approaches fail to capture authentic temporal reuse patterns of RAG workloads, resulting in a disconnect between theoretical optimization and real-world deployment efficiency.\par To bridge this gap, this paper introduces RAGPulse, an open-source RAG workload trace dataset that preserves authentic temporal information missing in existing traces. Collected from a production-grade university Q A system serving over 40,000 students and faculty members since April 2024, the dataset comprises 134K requests over 389 days. It unveils key RAG dynamics including time-sensitive document reuse, skewed access patterns over long time scales, and bursty temporal locality. By capturing these temporal reuse patterns, RAGPulse empowers the systems community to develop and validate next-generation optimization techniques, offering practical guidance for efficient RAG serving. The trace is available at https://github.com/flashserve/RAGPulse."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="wang-etal-2026-ragpulse">
<titleInfo>
<title>RAGPulse: A Real-World RAG Workload Capturing Authentic Temporal Reuse Patterns for Efficient Serving</title>
</titleInfo>
<name type="personal">
<namePart type="given">Zhengchao</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yitao</namePart>
<namePart type="family">Hu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zhuxuan</namePart>
<namePart type="family">Chang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jianing</namePart>
<namePart type="family">Ye</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Guotao</namePart>
<namePart type="family">Yang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tianhao</namePart>
<namePart type="family">Lai</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jiazheng</namePart>
<namePart type="family">Yu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zhixin</namePart>
<namePart type="family">Zhao</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Liang</namePart>
<namePart type="family">Zheng</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Deng</namePart>
<namePart type="family">Youpeng</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sheng</namePart>
<namePart type="family">Chen</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Keqiu</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-10</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: EMNLP 2026</title>
</titleInfo>
<name type="personal">
<namePart type="given">Hung-Yi</namePart>
<namePart type="family">Lee</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sunipa</namePart>
<namePart type="family">Dev</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jessy</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Desmond</namePart>
<namePart type="family">Elliott</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Budapest, Hungary</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-462-0</identifier>
</relatedItem>
<abstract>Retrieval-augmented generation (RAG) has emerged as a critical paradigm for deploying reliable and knowledge-intensive large language model (LLM) applications. Its unique execution pipeline introduces large overhead in system retrieval, such as massive context loading and document dependencies. To reduce this overhead, existing work on RAG serving systems often relies on generic LLM inference traces or synthetic RAG datasets. However, these approaches fail to capture authentic temporal reuse patterns of RAG workloads, resulting in a disconnect between theoretical optimization and real-world deployment efficiency.\par To bridge this gap, this paper introduces RAGPulse, an open-source RAG workload trace dataset that preserves authentic temporal information missing in existing traces. Collected from a production-grade university Q A system serving over 40,000 students and faculty members since April 2024, the dataset comprises 134K requests over 389 days. It unveils key RAG dynamics including time-sensitive document reuse, skewed access patterns over long time scales, and bursty temporal locality. By capturing these temporal reuse patterns, RAGPulse empowers the systems community to develop and validate next-generation optimization techniques, offering practical guidance for efficient RAG serving. The trace is available at https://github.com/flashserve/RAGPulse.</abstract>
<identifier type="citekey">wang-etal-2026-ragpulse</identifier>
<location>
<url>https://aclanthology.org/2026.findings-emnlp.335/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>7065</start>
<end>7076</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T RAGPulse: A Real-World RAG Workload Capturing Authentic Temporal Reuse Patterns for Efficient Serving
%A Wang, Zhengchao
%A Hu, Yitao
%A Chang, Zhuxuan
%A Ye, Jianing
%A Yang, Guotao
%A Lai, Tianhao
%A Yu, Jiazheng
%A Zhao, Zhixin
%A Zheng, Liang
%A Youpeng, Deng
%A Chen, Sheng
%A Li, Keqiu
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Findings of the Association for Computational Linguistics: EMNLP 2026
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-462-0
%F wang-etal-2026-ragpulse
%X Retrieval-augmented generation (RAG) has emerged as a critical paradigm for deploying reliable and knowledge-intensive large language model (LLM) applications. Its unique execution pipeline introduces large overhead in system retrieval, such as massive context loading and document dependencies. To reduce this overhead, existing work on RAG serving systems often relies on generic LLM inference traces or synthetic RAG datasets. However, these approaches fail to capture authentic temporal reuse patterns of RAG workloads, resulting in a disconnect between theoretical optimization and real-world deployment efficiency.\par To bridge this gap, this paper introduces RAGPulse, an open-source RAG workload trace dataset that preserves authentic temporal information missing in existing traces. Collected from a production-grade university Q A system serving over 40,000 students and faculty members since April 2024, the dataset comprises 134K requests over 389 days. It unveils key RAG dynamics including time-sensitive document reuse, skewed access patterns over long time scales, and bursty temporal locality. By capturing these temporal reuse patterns, RAGPulse empowers the systems community to develop and validate next-generation optimization techniques, offering practical guidance for efficient RAG serving. The trace is available at https://github.com/flashserve/RAGPulse.
%U https://aclanthology.org/2026.findings-emnlp.335/
%P 7065-7076
Markdown (Informal)
[RAGPulse: A Real-World RAG Workload Capturing Authentic Temporal Reuse Patterns for Efficient Serving](https://aclanthology.org/2026.findings-emnlp.335/) (Wang et al., Findings 2026)
ACL
- Zhengchao Wang, Yitao Hu, Zhuxuan Chang, Jianing Ye, Guotao Yang, Tianhao Lai, Jiazheng Yu, Zhixin Zhao, Liang Zheng, Deng Youpeng, Sheng Chen, and Keqiu Li. 2026. RAGPulse: A Real-World RAG Workload Capturing Authentic Temporal Reuse Patterns for Efficient Serving. In Findings of the Association for Computational Linguistics: EMNLP 2026, pages 7065–7076, Budapest, Hungary. Association for Computational Linguistics.