Database as Runtime: Compiling LLMs to SQL for In-database Model Serving
Summary: TranSQL+ compiles LLM inference to SQL for in-database serving, avoiding GPUs. Transforms transformer ops into relational algebra, yielding vector-oriented SQL that leverages DB buffers and indexes; demonstrated with LLaMA 3.1 8B on DuckDB. (summarized by gpt-5-nano on Feb 09 2026)
Incoming Non-self Citations Over Time
Authors
BibTeX Citation
@inproceedings{sun_sigmod25,
title = {{Database as Runtime: Compiling LLMs to SQL for In-database Model Serving}},
author = {Sun, Wenbo and Li, Ziyu and Hai, Rihan},
series = {{SIGMOD} '25},
booktitle = {Proceedings of the {ACM} {SIGMOD} International Conference on Management of Data},
publisher = {Association for Computing Machinery},
doi = {10.1145/3722212.3725093},
url = {https://dl.acm.org/doi/10.1145/3722212.3725093},
year = {2025}
}Incoming Citations (Sorted by Pagerank)
Showing 1 of 1 citing papers.
| Rank | Citing Paper | Year | Venue | Pagerank |
|---|---|---|---|---|
| 10,128 | Survivorship Bias in Industrial Database Workloads | 2026 | CIDR | 5.093636e-05 |
Previous
Page 1 / 1
Next
Outgoing Citations (Sorted by Pagerank)
Showing 7 of 7 cited papers.
Citations counted here include only citations to other VLDB/SIGMOD/CIDR/PODS papers in this database.
| Rank | Cited Paper | Year | Venue | Pagerank |
|---|---|---|---|---|
| 103 | DuckDB: an Embeddable Analytical Database | 2019 | SIGMOD | 0.00034161428 |
| 106 | The MADlib Analytics Library or MAD Skills, the SQL | 2012 | VLDB | 0.00033539462 |
| 536 | Learning Linear Regression Models over Factorized Joins | 2016 | SIGMOD | 0.0001693369 |
| 2,370 | Declarative Recursive Computation on an RDBMS or, Why You Should Use a Database For Distributed Machine Learning | 2019 | VLDB | 8.6795925e-05 |
| 3,653 | ClickHouse - Lightning Fast Analytics for Everyone | 2024 | VLDB | 7.2241143e-05 |
| 5,706 | SmartLite: A DBMS-based Serving System for DNN Inference in Resource-constrained Environments | 2024 | VLDB | 6.1144255e-05 |
| 6,046 | Automatic Optimization of Matrix Implementations for Distributed Machine Learning and Linear Algebra | 2021 | SIGMOD | 5.9956597e-05 |
Previous
Page 1 / 1
Next
Semantically Similar Papers
| # | Overall Rank | Paper | Year | Venue |
|---|---|---|---|---|
| 1 | 11,120 | Welding Natural Language Queries to Analytics IRs with LLMs | 2024 | CIDR |
| 2 | 279 | Text-to-SQL Empowered by Large Language Models: A Benchmark Evaluation | 2024 | VLDB |
| 3 | 4,944 | Hybrid Querying Over Relational Databases and Large Language Models | 2025 | CIDR |
| 4 | 10,510 | NL2SQLBench: A Modular Benchmarking Framework for LLM-Enabled NL2SQL Solutions | 2026 | VLDB |
| 5 | 7,809 | Can Large Language Models Be Query Optimizer for Relational Databases? | 2026 | SIGMOD |
| 6 | 7,685 | Beyond Quacking: Deep Integration of Language Models and RAG into DuckDB | 2025 | VLDB |
| 7 | 10,732 | RTS+: Reliable Text to SQL | 2025 | SIGMOD |
| 8 | 13,343 | Database Perspective on LLM Inference Systems | 2025 | VLDB |
| 9 | 4,045 | Logical and Physical Optimizations for SQL Query Execution over Large Language Models | 2025 | SIGMOD |
| 10 | 10,410 | TranSQL+: Serving Large Language Models with SQL on Low-Resource Hardware | 2026 | SIGMOD |