Horizon: Robust Checks for SQL Migration Using LLMs
Summary: Introduces Horizon, the first comprehensive LLM-based SQL schema migration framework that augments traditional DB tooling with novel syntactic and semantic checks to yield complete, functionally equivalent translations across all schema objects, including procedural code. Emphasizes robust evaluation and iterative query refinement to steer LLMs to high-accuracy migrations and provides configurable demos to probe strengths and limitations. (summarized by gpt-5-mini on Feb 09 2026)
Incoming Non-self Citations Over Time
No non-self incoming citations found for this paper in this database.
Authors
- 1. Venkatesh Emani (Microsoft)
- 2. Wenjing Wang (Microsoft)
- 3. Zi Ye (Microsoft)
- 4. Jia He (Microsoft)
- 5. Neel Ball (Microsoft)
- 6. Kumaraswamy Boora (Microsoft)
- 7. Carlo Curino (Microsoft)
- 8. Avrilia Floratou (Microsoft)
- 9. Manan Goenka (Microsoft)
- 10. Paridhi Gupta (Microsoft)
- 11. Vivek Gupta (Microsoft)
- 12. Katherine Lin (Microsoft)
- 13. Nick Litombe (Microsoft)
- 14. Jared Meade (Microsoft)
- 15. Suryakant Mutnal (Microsoft)
- 16. Mark Pryce-Maher (Microsoft)
- 17. Raghu Ramakrishnan (Microsoft)
- 18. Sudhir Raparla (Microsoft)
- 19. Dhruv Relwani (Microsoft)
- 20. Shyam Sai (Microsoft)
- 21. Vaibhave Sekar (Microsoft)
- 22. Roneet Shaw (Microsoft)
- 23. Harmeet Singh (Microsoft)
- 24. Prasanna Sridharan (Microsoft)
- 25. Mark Taylor (Microsoft)
- 26. Sunidhi Tiwari (Microsoft)
- 27. Yiwen Zhu (Microsoft)
BibTeX Citation
@article{emani_vldb25,
title = {{Horizon: Robust Checks for SQL Migration Using LLMs}},
author = {Emani, Venkatesh and Wang, Wenjing and Ye, Zi and He, Jia and Ball, Neel and Boora, Kumaraswamy and Curino, Carlo and Floratou, Avrilia and Goenka, Manan and Gupta, Paridhi and Gupta, Vivek and Lin, Katherine and Litombe, Nick and Meade, Jared and Mutnal, Suryakant and Pryce-Maher, Mark and Ramakrishnan, Raghu and Raparla, Sudhir and Relwani, Dhruv and Sai, Shyam and Sekar, Vaibhave and Shaw, Roneet and Singh, Harmeet and Sridharan, Prasanna and Taylor, Mark and Tiwari, Sunidhi and Zhu, Yiwen},
journal = {PVLDB},
series = {{VLDB} '25},
volume = {18},
number = {12},
pages = {5259--5262},
doi = {10.14778/3750601.3750646},
url = {https://doi.org/10.14778/3750601.3750646},
year = {2025}
}Incoming Citations (Sorted by Pagerank)
Showing 0 of 0 citing papers.
| Rank | Citing Paper | Year | Venue | Pagerank |
|---|
Outgoing Citations (Sorted by Pagerank)
Showing 2 of 2 cited papers.
Citations counted here include only citations to other VLDB/SIGMOD/CIDR/PODS papers in this database.
| Rank | Cited Paper | Year | Venue | Pagerank |
|---|---|---|---|---|
| 1,892 | Procedural Extensions of SQL: Understanding their usage in the wild | 2021 | VLDB | 9.5277793e-05 |
| 5,353 | QED: A Powerful Query Equivalence Decider for SQL | 2024 | VLDB | 6.2519268e-05 |
Semantically Similar Papers
| # | Overall Rank | Paper | Year | Venue |
|---|---|---|---|---|
| 1 | 5,744 | SQLStorm: Taking Database Benchmarking into the LLM Era | 2025 | VLDB |
| 2 | 5,323 | SNAILS: Schema Naming Assessments for Improved LLM-Based SQL Inference | 2025 | SIGMOD |
| 3 | 10,140 | BridgeScope: A Universal Toolkit for Bridging Large Language Models and Databases | 2026 | CIDR |
| 4 | 3,787 | Combining Small Language Models and Large Language Models for Zero-Shot NL2SQL | 2024 | VLDB |
| 5 | 2,602 | NL2SQL is a solved problem... Not! | 2024 | CIDR |
| 6 | 9,305 | The Power of Constraints in Natural Language to SQL Translation | 2025 | VLDB |
| 7 | 10,940 | Evoschema: Towards Text-To-Sql Robustness Against Schema Evolution | 2025 | VLDB |
| 8 | 10,139 | Leveraging Query Optimizers to Verify the Soundness of LLM-based Query Rewrites for Real-World Workloads, and More! | 2026 | CIDR |
| 9 | 6,936 | Reliable Text-to-SQL with Adaptive Abstention | 2025 | SIGMOD |
| 10 | 8,423 | Cracking SQL Barriers: An LLM-based Dialect Translation System | 2025 | SIGMOD |