Panda: Performance Debugging for Databases using LLM Agents
Summary: Panda: a framework that grounds pre-trained LLMs with database-specific runtime context and DBE-style workflows to produce in-context, actionable performance diagnoses and fixes. Key novelty: integrated Grounding, Verification, Affordance, and Feedback to ensure accuracy and deployability. (summarized by gpt-5-mini on Feb 09 2026)
Incoming Non-self Citations Over Time
Authors
- 1. Vikramank Singh (Amazon)
- 2. Kapil Eknath Vaidya (Amazon)
- 3. Vinayshekhar Bannihatti Kumar (Amazon)
- 4. Sopan Khosla (Amazon)
- 5. Balakrishnan Narayanaswamy (Amazon)
- 6. Rashmi Gangadharaiah (Amazon)
- 7. Tim Kraska (Amazon; Massachusetts Institute of Technology)
BibTeX Citation
@inproceedings{singh_cidr24,
address = {Amsterdam, Netherlands},
series = {{CIDR} '24},
title = {{Panda: Performance Debugging for Databases using LLM Agents}},
booktitle = {Proceedings of the {Conference} on {Innovative} {Data} {Systems} {Research}},
author = {Singh, Vikramank and Vaidya, Kapil Eknath and Kumar, Vinayshekhar Bannihatti and Khosla, Sopan and Narayanaswamy, Balakrishnan and Gangadharaiah, Rashmi and Kraska, Tim},
year = {2024}
}Incoming Citations (Sorted by Pagerank)
Showing 9 of 9 citing papers.
Previous
Page 1 / 1
Next
Outgoing Citations (Sorted by Pagerank)
Showing 6 of 6 cited papers.
Citations counted here include only citations to other VLDB/SIGMOD/CIDR/PODS papers in this database.
| Rank | Cited Paper | Year | Venue | Pagerank |
|---|---|---|---|---|
| 86 | Automatic Database Management System Tuning Through Large-scale Machine Learning | 2017 | SIGMOD | 0.00035316107 |
| 798 | DBSherlock: A Performance Diagnostic Tool for Transactional Databases | 2016 | SIGMOD | 0.00013919795 |
| 1,550 | Symphony: Towards Natural Language Query Answering over Multi-modal Data Lakes | 2023 | CIDR | 0.00010385904 |
| 1,949 | Diagnosing Root Causes of Intermittent Slow Queries in Cloud Databases | 2020 | VLDB | 9.430385e-05 |
| 3,586 | Make Your Database System Dream of Electric Sheep: Towards Self-Driving Operation | 2021 | VLDB | 7.2834069e-05 |
| 3,961 | MB2: Decomposed Behavior Modeling for Self-Driving Database Management Systems | 2021 | SIGMOD | 6.987575e-05 |
Previous
Page 1 / 1
Next