TAMO:Fine-Grained Root Cause Analysis via Tool-Assisted LLM Agent With Multi-Modality Observation Data in Cloud-Native Systems
Open accessThe DOI matches its Crossref record; 54 of the 54 references it cites were found in a registry or an index.
Abstract
Implementing large language models (LLMs)-driven root cause analysis (RCA) in cloud-native systems has become a key topic of modern software operations and maintenance. However, existing LLM-based approaches face three key challenges: multi-modality input constraint, context window limitation, and dynamic dependence graph. To address these issues, we propose a tool-assisted LLM agent with multi-modality observation data for fine-grained RCA, namely TAMO, including multi-modality alignment tool, root cause localization tool, and fault types classification tool. In detail, TAMO unifies multi-modal observation data into time-aligned representations for cross-modal feature consistency. Based on the unified representations, TAMO then invokes its specialized root cause localization tool and fault types classification tool for further identifying root cause and fault type underlying system context. This approach overcomes the limitations of LLMs in processing real-time raw observational data and dynamic service dependencies, guiding the model to generate repair strategies that align with system context through structured prompt design. Experiments on two benchmark datasets demonstrate that TAMO outperforms state-of-the-art (SOTA) approaches with comparable performance.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“the latest research positions them as reverse causal reasoning engines or reasoning hubs, realizing automated Root Cause Analysis (RCA) by scheduling external tools.”
From From Correlation to Intervention: A Survey on Log-Based Causal Analysis for Security and AIOps (2026) · cites this paper as [41]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“This division of labor alleviates the computational burden on LLMs, but lacks a feedback verification mechanism for intermediate results between the tools and the LLMs.”
From From Correlation to Intervention: A Survey on Log-Based Causal Analysis for Security and AIOps (2026) · cites this paper as [41]
Experimental — accuracy not yet measuredThe cited source says what the claim attributes to it.
In the citing paper
“TAMO [40] is a tool-assisted LLM-agent framework for fine-grained RCA that integrates multimodal alignment and model tools (e.g., localization/classification tools) to support root-cause analysis in cloud-native systems.”
From TopoEvo: A Topology-Aware Self-Evolving Multi-Agent Framework for Root Cause Analysis in Microservices (2026) · cites this paper as [40]
The cited source says what the claim attributes to it.
In the citing paper
“TAMO[41] proposes an innovative decoupled architecture: first utilizing diffusion models to fuse multimodal observational data, invoking frequency-domain and graph network tools to locate root cause nodes, and finally handing over the structured results to an LLM expert agent for comprehensive decision-making.”
From From Correlation to Intervention: A Survey on Log-Based Causal Analysis for Security and AIOps (2026) · cites this paper as [41]
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A. N., Kaiser, L., & Polosukhin, I. (2017) · arXiv · doi:10.48550/arxiv.1706.03762 · Cited by 120,404 · cited 3×
“Furthermore, since our research task is relatively novel and includes both fault localization and classification, we incorporate the same fault classifier used in TAMO for methods that do not natively implement fault classification (such as methods [23], [39], [42], etc.) for comparison.”
Adam: A Method for Stochastic Optimization
Kingma, D. P., & Ba, J. (2014) · arXiv · doi:10.48550/arxiv.1412.6980 · Cited by 98,946 · cited 1×
“The model is trained with the Adam optimizer [22], and all experiments use a fixed random seed.”
Exploiting Generative AI to Scale up Intelligent Tutoring Systems
Jakubův, J., Chvalovský, K., Goertzel, Z., Kaliszyk, C., Olšák, M., Piotrowski, B., Schulz, S., Suda, M., & Urban, J. (2023) · DROPS (Schloss Dagstuhl – Leibniz Center for Informatics) · doi:10.4230/lipics.itp.2023.19 · Cited by 76,658
DeepLog: Anomaly Detection and Diagnosis from System Logs through Deep Learning
Du, M., Li, F., Zheng, G., & Srikumar, V. (2017) · Proceedings of the 2017 ACM SIGSAC Conference on Computer and Communications Security · doi:10.1145/3133956.3134015 · Cited by 1,835 · cited 1×
“DeepLog[11] modelslogfeaturesvialongshort-termmemorynetworks.”
Graph Neural Network-Based Anomaly Detection in Multivariate Time Series
Deng, A., & Hooi, B. (2021) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v35i5.16523 · Cited by 1,413 · cited 1×
“GDN[10]andGTA[8]constructgraphneuralnetworksusing metricsdatatocapturefaultpropagationpaths.”
Drain: an online log parsing approach with fixed depth tree
He, P., Zhu, J., Zheng, Z., & Lyu, M. R. (2017) · 2017 IEEE International Conference on Web Services (ICWS) · doi:10.1109/icws.2017.13 · Cited by 943 · cited 1×
“…are difficult to extract high-quality features, inspired by previous work [14], [23], we extract event patterns E and keywords W from the raw logs using Drain [16], and by calculating the TF-IDF scores, we select the highest scoring pattern-keyword pairs as the conditional variable c log .”
How Attentive are Graph Attention Networks?
Brody, S., Alon, U., & Yahav, E. (2021) · arXiv · doi:10.48550/arxiv.2105.14491 · Cited by 789
Burns, B., Grant, B., Oppenheimer, D., Brewer, E., & Wilkes, J. (2016) · Communications of the ACM · doi:10.1145/2890784 · Cited by 690 · cited 1×
“Among these, 9 fault types are related to services and pods in the context of Kubernetes (K8s) containers [4], while node faults include 6 types: sudden memory pressure, disk space exhaustion, disk read/write issues, CPU pressure, and slow CPU growth.”
Large language models for software engineering: survey and open problems
Fan, A., Gokkaya, B., Harman, M., Lyubarskiy, M., Sengupta, S., Yoo, S., & Zhang, J. M. (2023) · 2023 IEEE/ACM International Conference on Software Engineering: Future of Software Engineering (ICSE-FoSE) · doi:10.1109/icse-fose59343.2023.00008 · Cited by 650 · cited 1×
“Recently, large language models (LLMs) have demonstrated revolutionary capabilities in automated RCA, opening new avenues for AIOps [9], [12].”
Learning Graph Structures With Transformer for Multivariate Time-Series Anomaly Detection in IoT
Chen, Z., Chen, D., Zhang, X., Yuan, Z., & Cheng, X. (2022) · IEEE Internet of Things Journal · doi:10.1109/jiot.2021.3100509 · Cited by 595 · cited 1×
“GDN[10]andGTA[8]constructgraphneuralnetworksusing metricsdatatocapturefaultpropagationpaths.”
Self-citation
Research works printed but not linked: 3
- Ho, Denoising diffusion probabilistic models, 6840
- Ke, LightGBM: A highly efficient gradient boosting decision tree, 3149
- Sarda, ADARMA auto-detection and auto-remediation of microservice anomalies by leveraging large language models, 200
Cited by
ThinkFL: Self-Refining Failure Localization for Microservice Systems via Reinforcement Fine-Tuning
Zhang, L., Zhai, Y., Jia, T., Duan, C., Yu, S., Gao, J., Ding, B., Wu, Z., & Li, Y. (2026) · ACM Transactions on Software Engineering and Methodology · doi:10.1145/3789262 · Cited by 31
Agentic Memory Enhanced Recursive Reasoning for Root Cause Localization in Microservices
Zhang, L.-Z., Jia, T., Zhai, Y.-P., Pan, L.-Y., Duan, C.-M., He, M.-H., Jia, M., & Li, Y. (2026) · SEIP@ICSE · doi:10.1145/3786583.3786853 · Cited by 16
Hypothesize-Then-Verify: Speculative Root Cause Analysis for Microservices with Pathwise Parallelism
Zhang, L., Jia, T., Zhai, Y., Pan, L., Duan, C., He, M., Xiao, P., & Li, Y. (2026) · ACM/IEEE International Conference on Software Engineering: Companion Proceedings (ICSE-Companion) · doi:10.1145/3786582.3786803 · Cited by 17
“…and reasoning capabilities, LLMs can reduce dependence on handcrafted features while offering strong cross-platform adaptability, task generalization, and flexibility, thereby alleviating the limitations of conventional methods [2, 4, 8–10, 12, 13, 17, 18, 21, 24, 28, 31, 34, 35, 38, 39].”
Riddell, E., Riddell, J., Sun, G., Antkiewicz, M., & Czarnecki, K. (2026) · IEEE/ACM International Conference on AI Foundation Models and Software Engineering (FORGE) · doi:10.1145/3793655.3793732 · Cited by 15
“More recent work integrates LLMs as agents into retrieval-or tool-augmented frameworks [25, 28, 50] or multi-agent systems [14, 21, 35, 40, 48].”
Tang, Y., Chen, K., Yue, L., Fan, J., Zhou, C., Li, X., Zhang, Y., Zhao, M., Kai, S., Guo, K., et al. (2025) · arXiv (Cornell University) · doi:10.48550/arxiv.2510.17491 · Cited by 10
“Frameworks like RCAgent [262] and TAMO [263] leverage tool-enhanced LLMs to perform root cause analysis for industrial systems or microservices architectures.”
Metadata
- Authors
- Zhang, X., Wang, Q., Li, M., Yuan, Y., Xiao, M., Zhuang, F., & Yu, D.
- Affiliations
- Shandong University (CN)
- Beihang University (CN)
- Venue
- IEEE Transactions on Services Computing
- Published
- 2025-11
- Volume
- 18
- Issue
- 6
- Pages
- 4221-4233
- ISSN
- 1939-1374 (ISSN-L)
- 2372-0204
- OpenAlex
- W4415933773
- Cited by
- 15
- References
- 38
- Citations by year
- 2026: 15
- FWCI
- 6.22
- Type
- journal article
- Language
- en
- Open access
- green
- arxiv.org
- Licence
- Locations
- IEEE Transactions on Services Computing (publishedVersion)
- arXiv (Cornell University) (submittedVersion)
- arXiv (Cornell University)
- Funders
- National Natural Science Foundation of China (62202273, 62176014)
- Joint Key Funds of National Natural Science Foundation of China (U24B20149)
- Major Basic Research Program of Shandong Provincial Natural Science Foundation (ZR2025ZD18)
- China Postdoctoral Science Foundation (2024M761806)
- Fundamental Research Funds for the Central Universities
- Topics
- Physical Sciences › Computer Science › Computer Networks and Communications › Software System Performance and Reliability
- Physical Sciences › Computer Science › Information Systems › Software Engineering Research
- Physical Sciences › Computer Science › Information Systems › Cloud Computing and Resource Management
- Keywords
- root cause analysis, large language models, LLM agents, cloud-native systems, root cause localization, fault classification, cross-modal alignment, service dependency graph, repair strategies, prompt engineering
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: TAMO:Fine-Grained Root Cause Analysis via Tool-Assisted LLM Agent With Multi-Modality Observation Data in Cloud-Native Systems. https://app.cochat.ai/papers/10.1109/tsc.2025.3629066
- The paper
- Zhang, X., Wang, Q., Li, M., Yuan, Y., Xiao, M., Zhuang, F., & Yu, D. (2025). TAMO:Fine-Grained Root Cause Analysis via Tool-Assisted LLM Agent With Multi-Modality Observation Data in Cloud-Native Systems. IEEE Transactions on Services Computing. https://doi.org/10.1109/tsc.2025.3629066
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.