Bespoke OLAP: Synthesizing Workload-Specific One-size-fits-one Database Engines

Summary: Bespoke OLAP autonomously synthesizes workload-specific OLAP engines via iterative performance evaluation and validation. Generated storage layouts and algorithms avoid generality overhead, yielding order-of-magnitude speedups over DuckDB and Umbra. (summarized by gpt-5.6-luna on Aug 28 2026)

Paper ID: hd8a25f9a341761f5
Venue: VLDB
Year: 2026
Pagerank: 5.4255843e-05
Overall Rank: 7,844 | 47.43%
DOI: 10.14778/3836663.3836723
PDF: Download (CC BY-NC-ND 4.0)

Incoming Non-self Citations Over Time

Authors

1. Johannes Wehrstein (Technical University of Darmstadt)
2. Timo Eckmann (Technical University of Darmstadt)
3. Matthias Jasny (Microsoft)
4. Carsten Binnig (German National Research Center for Information Technology; Hessian Center for Artificial Intelligence; Technical University of Darmstadt)

BibTeX Citation

@article{wehrstein_vldb26,
        title = {{Bespoke OLAP: Synthesizing Workload-Specific One-size-fits-one Database Engines}},
        author = {Wehrstein, Johannes and Eckmann, Timo and Jasny, Matthias and Binnig, Carsten},
        journal = {PVLDB},
        series = {{VLDB} '26},
        volume = {19},
        number = {11},
        pages = {3759--3771},
        doi = {10.14778/3836663.3836723},
        url = {https://doi.org/10.14778/3836663.3836723},
        year = {2026}
}

Incoming Citations (Sorted by Pagerank)

Showing 2 of 2 citing papers.

Rank	Citing Paper	Year	Venue	Pagerank
10,893	Decisionhouse: Prescriptive Analytics in the Data Stack	2026	VLDB	4.9605343e-05
11,043	Demonstrating GenDB: Instance-Optimized and Customized Query Processing Code Generation via LLM Agents	2026	VLDB	4.9605343e-05

Outgoing Citations (Sorted by Pagerank)

Showing 22 of 22 cited papers.

Citations counted here include only citations to other VLDB/SIGMOD/CIDR/PODS papers in this database.

Rank	Cited Paper	Year	Venue	Pagerank
14	MonetDB/X100: Hyper-Pipelining Query Execution	2005	CIDR	0.00063972112
15	How Good Are Query Optimizers, Really?	2016	VLDB	0.00061463175
21	Efficiently Compiling Efficient Query Plans for Modern Hardware	2011	VLDB	0.00056737487
40	The Case for Learned Index Structures	2018	SIGMOD	0.00046249672
74	Automatic Database Management System Tuning Through Large-scale Machine Learning	2017	SIGMOD	0.00037164008
84	Learned Cardinalities: Estimating Correlated Joins with Deep Learning	2019	CIDR	0.00035920657
89	Automated Selection of Materialized Views and Indexes for SQL Databases	2000	VLDB	0.00035265687
142	Neo: A Learned Query Optimizer	2019	VLDB	0.00029114937
151	An Efficient, Cost-Driven Index Selection Tool for Microsoft SQL Server	1997	VLDB	0.00028611216
196	Integrating Vertical and Horizontal Partitioning into Automated Physical Database Design	2004	SIGMOD	0.00025581936
224	Self-Driving Database Management Systems	2017	CIDR	0.00023973185
375	Umbra: A Disk-Based System with In-Memory Performance	2020	CIDR	0.00019660459
380	AutoAdmin "What-if" Index Analysis Utility	1998	SIGMOD	0.00019530546
408	DBToaster: Higher-order Delta Processing for Dynamic, Frequently Fresh Views	2012	VLDB	0.00018836656
493	Database Tuning Advisor for Microsoft SQL Server 2005	2004	VLDB	0.0001735831
498	Building Efficient Query Engines in a High-Level Language	2014	VLDB	0.00017323427
1,740	Flow-Loss: Learning Cardinality Estimates That Matter	2021	VLDB	9.7289523e-05
1,791	Why TPC Is Not Enough: An Analysis of the Amazon Redshift Fleet	2024	VLDB	9.6212919e-05
1,931	CodexDB: Synthesizing Code for Query Processing from Natural Language Instructions using GPT-3 Codex	2022	VLDB	9.3449634e-05
6,085	Demonstrating GPT-DB: Generating Query-Specific and Customizable Code for SQL Processing with GPT-4	2023	VLDB	5.8745517e-05
10,258	DBMS Fitting: Why should we learn what we already know?	2020	CIDR	5.0391159e-05
10,278	Redbench: Workload Synthesis From Cloud Traces	2026	VLDB	5.0334833e-05

Semantically Similar Papers

#	Overall Rank	Paper	Year	Venue
1	8,701	Adaptive Code Generation for Data-Intensive Analytics	2021	VLDB
2	6,095	From Auto-tuning One Size Fits All to Self-designed and Learned Data-intensive Systems	2019	SIGMOD
3	10,371	QueryArtisan: Generating Data Manipulation Codes for Ad-hoc Analysis in Data Lakes	2025	VLDB
4	11,449	A Demonstration of QueryArtisan: Real-Time Data Lake Analysis via Dynamically Generated Data Manipulation Code	2025	VLDB
5	9,494	SQLBarber: A System Leveraging Large Language Models to Generate Customized and Realistic SQL Workloads	2026	SIGMOD
6	4,061	Database-Agnostic Workload Management	2019	CIDR
7	8,849	Boosting Efficiency of External Pipelines by Blurring Application Boundaries	2022	CIDR
8	5,211	Database Workload Characterization with Query Plan Encoders	2022	VLDB
9	11,043	Demonstrating GenDB: Instance-Optimized and Customized Query Processing Code Generation via LLM Agents	2026	VLDB
10	10,323	This is Going to Sound Crazy, But What If We Used Large Language Models to Boost Automatic Database Tuning Algorithms By Leveraging Prior History? We Will Find Better Configurations More Quickly Than Retraining From Scratch!	2026	SIGMOD