ORCA: Observability-Grounded Program Repair for Microservice Incidents
Source
arxiv.org
Author
Yuanchen Gao, Yifang Tian, Yiran Li, Charles Zhang, Hans-Arno Jacobsen
Date
Why it matters
groundingTying a model's answers to checkable sources — retrieved documents, live data, tool results — instead of letting it answer from memory alone.Full definition → an AI agentAn AI system that doesn't just answer once but works toward a goal in a loop — taking actions, reading the results, and deciding what to do next.Full definition → in the operational signal you actually have during an incident, rather than a tidy issue report, is the more honest version of automated repair.
Terms in this piece · Glossary
grounding — Tying a model's answers to checkable sources — retrieved documents, live data, tool results — instead of letting it answer from memory alone.
AI agent — An AI system that doesn't just answer once but works toward a goal in a loop — taking actions, reading the results, and deciding what to do next.
articleRefine After Generation: Toward Correct and Concise Patches in LLM-based Program RepairWenqiang Luo, Jacky Keung, Xiaoyu Shi, Yicheng Sun, Boyang Yang, Zhou Yang, Haoye Tian
articleAdaRepair-Mem: Adaptive Experience Orchestration for Repository-Level Program RepairZ. C. Luo, J. C. Guo, W. J. He, S. Y. Wang, J. C. Yu, F. M. Zhao, Y. Chen, T. Cao, L. Q. Liu, N. Zheng, W. Xu, J. Jiang, Z. M. Zhao
articleSupporting Industrial Test-Failure Analysis with LLM-Based Systems: An Experience ReportEric Jansson, Per Strandberg, Thomas S\"orensen, Eduard Paul Enoiu, Wasif Afzal