DeepEye-SQL: A Software-Engineering-Inspired Text-to-SQL Framework DOI: XXXXXXX.XXXXXXX
Boyan Li, Chong Chen, Zhujun Xue, Yinan Mei, Yuyu Luo · 2025
Read DeepEye-SQL: A Software-Engineering-Inspired Text-to-SQL Framework DOI: XXXXXXX.XXXXXXX by Boyan Li, Chong Chen, Zhujun Xue, Yinan Mei, Yuyu Luo (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2510-17586.txt.
Description
Boyan Li, with Chong Chen and three co-authors, treats database questions as software projects in a 2025 arXiv paper. Their system, DeepEye-SQL, turns plain language questions into database code. Single model calls often produce code that looks right but fails when run. The authors argue the real gap is weak orchestration, meaning poor control over the full workflow, not weak parts.
Four stages mirror the classic software development life cycle. First, the system links question words to the right tables and columns. Next, it drafts several code versions for safety. A checking chain then screens syntax, logic, and quality before anything runs. Finally, run results guide the choice among drafts. Without extra training, the system beat larger rivals on standard benchmarks.
Catalog record
| CREATOR | Boyan Li, Chong Chen, Zhujun Xue, Yinan Mei, Yuyu Luo |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 67 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2510-17586