Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Zhu, Yiwen, Cahoon, Joyce, Pavlenko, Anna, Bai, Qiushi, Shahbazi, Nima, Vermareddy, Divya, Wang, Meina, Demarne, Mathieu, Bararia, Swati, Wang, Wenjing, Kumar, Hemkesh Vijaya, Lerner, Hannah, Lin, Katherine, Toscano, Steve, Cilimdzic, Miso, Krishnan, Subru
Format: Preprint
Veröffentlicht: 2026
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2605.17617
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866910256629547008
author Zhu, Yiwen
Cahoon, Joyce
Pavlenko, Anna
Bai, Qiushi
Shahbazi, Nima
Vermareddy, Divya
Wang, Meina
Demarne, Mathieu
Bararia, Swati
Wang, Wenjing
Kumar, Hemkesh Vijaya
Lerner, Hannah
Lin, Katherine
Toscano, Steve
Cilimdzic, Miso
Krishnan, Subru
author_facet Zhu, Yiwen
Cahoon, Joyce
Pavlenko, Anna
Bai, Qiushi
Shahbazi, Nima
Vermareddy, Divya
Wang, Meina
Demarne, Mathieu
Bararia, Swati
Wang, Wenjing
Kumar, Hemkesh Vijaya
Lerner, Hannah
Lin, Katherine
Toscano, Steve
Cilimdzic, Miso
Krishnan, Subru
contents Complex operational workflows coordinating personnel, tools, and information are central to system operations, yet end-to-end automation remains challenging due to extensive human input requirements and limited ability to adapt over time. We present GraphMind, a system that constructs, executes, and evolves action-centric workflow graphs with minimal human effort. The system operates in three phases. First, a scalable offline pipeline extracts structured workflow graphs from large volumes of human resolution traces, capturing problems, actions, and their causal relationships. Second, an online multi-agent traversal engine navigates the graph to dynamically construct and execute workflows, combining graph-guided retrieval with LLM-driven reasoning at each step. Third, Adaptive Traversal Reinforcement (ATR) reinforces successful traversal paths, enabling execution-informed graph adaptation. GraphMind has been deployed across four production cloud database services for incident investigation. Evaluated on 93 held-out incidents and validated via blind expert review, the system outperforms an Agentic Summary-RAG baseline in mitigation reach, hallucination rate, and diagnostic throughput while requiring 8x less retrieval context. The ATR layer reduces hallucination rate by 26%, demonstrating that workflow graphs can learn from execution feedback. A 12-week field study confirms practical value: 97% of scored conversations yield actionable results within interactive latency.
format Preprint
id arxiv_https___arxiv_org_abs_2605_17617
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle GraphMind: From Operational Traces to Self-Evolving Workflow Automation
Zhu, Yiwen
Cahoon, Joyce
Pavlenko, Anna
Bai, Qiushi
Shahbazi, Nima
Vermareddy, Divya
Wang, Meina
Demarne, Mathieu
Bararia, Swati
Wang, Wenjing
Kumar, Hemkesh Vijaya
Lerner, Hannah
Lin, Katherine
Toscano, Steve
Cilimdzic, Miso
Krishnan, Subru
Artificial Intelligence
Complex operational workflows coordinating personnel, tools, and information are central to system operations, yet end-to-end automation remains challenging due to extensive human input requirements and limited ability to adapt over time. We present GraphMind, a system that constructs, executes, and evolves action-centric workflow graphs with minimal human effort. The system operates in three phases. First, a scalable offline pipeline extracts structured workflow graphs from large volumes of human resolution traces, capturing problems, actions, and their causal relationships. Second, an online multi-agent traversal engine navigates the graph to dynamically construct and execute workflows, combining graph-guided retrieval with LLM-driven reasoning at each step. Third, Adaptive Traversal Reinforcement (ATR) reinforces successful traversal paths, enabling execution-informed graph adaptation. GraphMind has been deployed across four production cloud database services for incident investigation. Evaluated on 93 held-out incidents and validated via blind expert review, the system outperforms an Agentic Summary-RAG baseline in mitigation reach, hallucination rate, and diagnostic throughput while requiring 8x less retrieval context. The ATR layer reduces hallucination rate by 26%, demonstrating that workflow graphs can learn from execution feedback. A 12-week field study confirms practical value: 97% of scored conversations yield actionable results within interactive latency.
title GraphMind: From Operational Traces to Self-Evolving Workflow Automation
topic Artificial Intelligence
url https://arxiv.org/abs/2605.17617