At the main conference, undergraduate alumni Daniel Xue presented his paper about fully concurrent GROUP BY aggregation, which was his senior thesis at Penn last year. Daniel’s work shows that a simple custom hash table may be the key ingredient!

At the AIDB workshop, Ph.D. student Jeff Tao presented a full paper on generating query optimization benchmarks with Bayesian optimization via Gremlin, which won a best paper honorable mention! Jeff’s work produced queries with an order of magnitude more headroom (optimization potential) than existing, hand-crafted benchmarks.

Also the AIDB workshop, MA/MS graduate Wangshu Hong presented a short paper about extending learned query optimizers, like Balsa, to DuckDB. Wangshu’s work also showed how censored observations could improve training stability for RL-powered learned optimizers.

At DASHSys, Ph.D. student Zixuan Yi presented her work on Data Canvas, a provenance-guided harness for agentic data engineering. Zixuan’s work allows a swarm of agents to collaboratively build large datasets, with hard lineage guarantees and support for human-driven repairs. A full version of Zixuan’s work will be presented at COLM this year.

The paper argues that workload traces observed in industrial settings represent a negotiation between the data platform and the platform’s users. Users mold their queries to run well on the platform, and engineers tune the platform to better meet user demands. This cycle, while great for both users and data platforms, creates a survivorship bias in the observed workloads: the most frequently processed queries are precisely the queries that are already working well.

The paper’s authors include Ryan Marcus (assistant professor), Jeffrey Tao (PhD student), Peizhi Wu (PhD student alumnus), and Zijie Zhao (PhD student).

Penn students presented five posters:

• Zhen Ping Khor presented “Adaptive load balancing in DAG-based consensus protocols” (link)
• Zijie Zhao presented “Low-Overhead Adaptive Execution via Reinforcement Learning with Bootstrapped Counterfactuals”
• Jeff Tao presented “Adversarial Benchmark Generation” (link)
• Phillip Hilliard presented “Accurate ML Processing under Real-Time Constraints” (link)
• Varun Jana presented “Semantic Indexing for Unstructured Data Management at Scale”

In addition to student posters, Ryan presented one of the three keynotes, “Next Steps for Learned Query Optimization” (slides).

Ph.D. student Soonbo Han (advisor: Zachary Ives) presented his work titled “Implementation Strategies for Views over Property Graphs,” which won the best paper award! Soonbo’s work shows query rewriting techniques can take advantage of semantic views over graph data, including how to index and maintain such views dynamically.

Ph.D. student Jiaming Liang (advisor: Zachary Ives) presented his work titled “RITA: Group Attention is All You Need for Timeseries Analytics.” Jiaming’s work shows how careful grouping and caching of semantically-similar inputs can accelerate neural attention mechanisms, allowing attention networks to scale up to previously-impossible tasks.

Due to visa issues, Zack presented a paper from Ph.D. student Peizhi Wu, titled “Modeling Shifting Workloads for Learned Database Systems.” Peizhi’s work shows how to keep learned database components up to date with data drift using a carefully-tuned replay buffer.

Visiting Ph.D. student Yao Tian (advisor: Xiaofang Zhou, Penn supervisors: Zachary Ives and Ryan Marcus) presented her work titled “A Learned Cuckoo Filter for Approximate Membership Queries over Variable-sized Sliding Windows on Data Streams.” Yao’s work combines traditional Cuckoo filters with deep learning models to enable approximate membership queries over dynamically-sized windows, achieving significantly higher accuracy than previous results.

At the aiDM workshop, Ph.D. student Zixuan Yi (advisor: Ryan Marcus and Zachary Ives) presented her work titled “Low Rank Approximation for Learned Query Optimization.” Zixuan’s work shows how linear methods for approximating low rank matrices can be used to learn to steer an entire query workload at once.

Later this summer, Penn will present several papers at VLDB in Guangzhou, China!

Ph.D. student Jeff Tao (supervisors: Ryan Marcus and Andrew Head) presented his work on using Bayesian optimization and deep learning to search for optimal query plans. This search process can be considered “superoptimization,” since the search process itself executes query plans. Jeff’s work has the potential to find gains in highly repetitive and costly analytic queries.

Ph.D. student Zixuan Yi (supervisors: Zachary Ives and Ryan Marcus) presented her work on workload-level query steering. Zixuan showed that the space of optimizer configurations for each query in a workload can be effectively explored using matrix completion. Zixuan’s work is also described in her recent workshop paper (PDF).

Both Jeff and Zixuan received valuable feedback at the event, and plan to turn their work into full papers soon. Undergraduate Austin Wang, visiting Ph.D. student Yao Tian, and Ryan Marcus also came along!

You can check out Zixuan and Jeff’s posters below:

Ph.D. student Chenyuan Wu (supervisor: Boon Thau Loo) presented work on FlexChain and AdaChain. FlexChain (PDF) is the first work to present a blockchain that disaggregates compute, storage, RAM, and cold storage to better serve diverse workloads. AdaChain (PDF) shows how reinforcement learning can be used to adaptively switch a blockchain system between different architectures, keeping up with workload or hardware changes.

AutoSteer (PDF), an extensible system for steering the query optimizer of any SQL database, was presented by collaborator and TUM Ph.D. student Christoph Anneser. AutoSteer is open source, and has been tested on Meta’s Presto cluster. The AutoSteer work was a collaboration between UPenn, Intel Labs, MIT, Meta, and TU Munich.

Ph.D. student Bhavana Mehta (supervisor: Boon Thau Loo) presented RLShard (PDF), a vision for a Byzantine fault tolerant, sharded, transactional database at the aiDB workshop. RLShard contains designs for both a fully decentralized, trustless system, as well as a semi-centralized version taking advantage of an administrative domain. Bhavana will continue developing both approaches as part of her dissertation.

Also at aiDB, assistant professor Ryan Marcus presented a vision for Learned Query Superoptimization (PDF), a set of new techniques to improve common repetitive queries in analytic database systems. Ryan plans to implement this vision in the coming years.

The Penn DB Group will see you at the next one!

• Jeff Tao, who previously earned his master’s degree from Columbia University, is joining the lab to work on data visualization and analysis systems that are easy to use and optimized end-to-end.
• Zixuan Yi, who previously earned her bachelor’s degree from Tsinghua University, is joining the lab to work on query optimization and learned systems.