ML for Systems

This project aims to develop a learning-based distributed transaction processing system that can be deployed in untrusted environments. We argue that as novel transactional logics, modern hardware, and new cloud platforms arise, distributed transaction processing systems need to be designed withfull-stack adaptivityin mind. At the application level, such a system must adaptively learn the best-performing transaction processing paradigm and quickly adapt to new hardware and unanticipated workload changes on the fly. Likewise, the Byzantine consensus layer must dynamically adjust itself to the workloads, faulty conditions, and network configuration while maintaining compatibility with the transaction processing paradigm. At the infrastructure level, cloud providers must enable cross-layer adaptation, which identifies performance bottlenecks and possible attacks, and determines at runtime the degree of resource disaggregation that best meets application requirements.

Publications

1. Chenyuan Wu, Mohammad Javad Amiri, Jared Asch, Heena Nagda, Qizhen Zhang, and Boon Thau Loo. “FlexChain: an elastic disaggregated blockchain.” Proceedings of the VLDB Endowment, 2023.

2. Chenyuan Wu, Bhavana Mehta, Mohammad Javad Amiri, Ryan Marcus, and Boon Thau Loo. “AdaChain: A Learned Adaptive Blockchain.” Proceedings of the VLDB Endowment, 2023.

3. Chenyuan Wu, Haoyun Qin, Mohammad Javad Amiri, Boon Thau Loo, Dahlia Malkhi, and Ryan Marcus. “BFTBrain: Adaptive BFT Consensus with Reinforcement Learning.” USENIX Symposium on Networked Systems Design and Implementation (NSDI), 2025.

4. Chenyuan Wu, Mohammad Javad Amiri, Haoyun Qin, Bhavana Mehta, Ryan Marcus, and Boon Thau Loo. “Towards Full Stack Adaptivity in Permissioned Blockchains.” Proceedings of the VLDB Endowment, 2024.

5. Bhavana Mehta, Nupur Baghel, Mohammad Javad Amiri, Boon Thau Loo, and Ryan Marcus. “Adaptive Sharding in Untrusted Environments.” Proceedings of the ACM on Management of Data (SIGMOD), 2026.

6. Chenyuan Wu, Haoyun Qin, Mohammad Javad Amiri, Boon Thau Loo, Dahlia Malkhi, and Ryan Marcus. “Towards truly adaptive byzantine fault-tolerant consensus.” ACM SIGOPS Operating Systems Review 58, 2024.

7. Bhavana Mehta , Neelesh C. A, Prashanth S. Iyer, Mohammad Javad Amiri, Boon Thau Loo, Ryan Marcus. “Towards adaptive fault-tolerant sharded databases (Extended Abstracts).” Workshop on Applied AI for Database Systems and Applications(AIDB@VLDB’23)

People

Eric Zhou

Simons Research Fellow

Chenyuan Wu

Assistant Professor, City University of Hong Kong

Haoyun Qin

Research Scientist at Spellbrush

Bhavana Mehta

PhD, Penn, 2025 (co-advised with Boon), next: MTS, Zyphra

Ryan Marcus

Assistant Professor, University of Pennsylvania

Boon Thau Loo

Professor, University of Pennsylvania

Mohammad Javad Amiri

Assistant Professor at Stony Brook University