PRISM: Navigating Cost–Accuracy Trade-offs for NL2SQL
Summary: PRISM targets NL2SQL deployment as a schema-aware cost–accuracy optimization problem, rather than pure accuracy maximization. Uses cost-aware Bayesian optimization offline to curate high-performing pipelines, then deploys a single candidate or ensemble for strong accuracy at dramatically lower inference cost. (summarized by gpt-5-mini on Apr 11 2026)
Incoming Non-self Citations Over Time
No non-self incoming citations found for this paper in this database.
Authors
- 1. Gaurav Tarlok Kakkar (Georgia Institute of Technology)
- 2. Yeounoh Chung (Google)
- 3. Fatma Özcan (Google)
- 4. Steve Mussmann (Georgia Institute of Technology)
- 5. Joy Arulraj (Georgia Institute of Technology)
BibTeX Citation
@inproceedings{kakkar_sigmod26,
title = {{PRISM: Navigating Cost–Accuracy Trade-offs for NL2SQL}},
author = {Kakkar, Gaurav Tarlok and Chung, Yeounoh and Özcan, Fatma and Mussmann, Steve and Arulraj, Joy},
series = {{SIGMOD} '26},
booktitle = {Proceedings of the {ACM} {SIGMOD} International Conference on Management of Data},
publisher = {Association for Computing Machinery},
doi = {10.1145/3786679},
url = {https://dl.acm.org/doi/10.1145/3786679},
year = {2026}
}Incoming Citations (Sorted by Pagerank)
Showing 0 of 0 citing papers.
| Rank | Citing Paper | Year | Venue | Pagerank |
|---|
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 |
|---|---|---|---|---|
| 174 | Text-to-SQL Empowered by Large Language Models: A Benchmark Evaluation | 2024 | VLDB | 0.00026790979 |
| 533 | CodeS: Towards Building Open-source Language Models for Text-to-SQL | 2024 | SIGMOD | 0.00016826571 |
| 683 | DocETL: Agentic Query Rewriting and Evaluation for Complex Document Processing | 2025 | VLDB | 0.00014817539 |
| 748 | Palimpzest: Optimizing AI-Powered Analytics with Declarative Query Processing | 2025 | CIDR | 0.00014281926 |
| 1,777 | OmniSQL: Synthesizing High-quality Text-to-SQL Data at Scale | 2025 | VLDB | 9.664552e-05 |
| 5,116 | Is Long Context All You Need? Leveraging LLM's Extended Context for NL2SQL | 2025 | VLDB | 6.2678524e-05 |
| 6,860 | SpareLLM: Automatically Selecting Task-Specific Minimum-Cost Large Language Models under Equivalence Constraint | 2025 | SIGMOD | 5.6613762e-05 |
Previous
Page 1 / 1
Next
Semantically Similar Papers
| # | Overall Rank | Paper | Year | Venue |
|---|---|---|---|---|
| 1 | 174 | Text-to-SQL Empowered by Large Language Models: A Benchmark Evaluation | 2024 | VLDB |
| 2 | 7,012 | Reliable Text-to-SQL with Adaptive Abstention | 2025 | SIGMOD |
| 3 | 4,924 | SNAILS: Schema Naming Assessments for Improved LLM-Based SQL Inference | 2025 | SIGMOD |
| 4 | 2,348 | The Dawn of Natural Language to SQL: Are We Fully Ready? | 2024 | VLDB |
| 5 | 2,117 | NL2SQL is a solved problem... Not! | 2024 | CIDR |
| 6 | 10,494 | Prism: Private Relational Data Synthesis with Language Models | 2026 | SIGMOD |
| 7 | 7,952 | The Power of Constraints in Natural Language to SQL Translation | 2025 | VLDB |
| 8 | 3,489 | Combining Small Language Models and Large Language Models for Zero-Shot NL2SQL | 2024 | VLDB |
| 9 | 10,766 | A Comparative Evaluation of Schema Subsetting for LLM-based NL-to-SQL over Large-Schema Databases | 2026 | VLDB |
| 10 | 10,695 | NL2SQLBench: A Modular Benchmarking Framework for LLM-Enabled NL2SQL Solutions | 2026 | VLDB |