Chunwei Liu - Homepage

Short Bio

Chunwei Liu is a Postdoctoral Associate at MIT CSAIL working with Michael Cafarella. He got his Ph.D. from Department of Computer Science at the University of Chicago, where he worked in ChiData group advised by Aaron Elmore. His research interests span compound AI systems, database systems, cloud/edge computing, and database benchmarking. He focuses on optimizing data systems for both conventional data analytics and emerging AI-powered pipelines. Chunwei develops privacy-preserving workload generation techniques for evaluating cloud database systems, collaborating with major cloud vendors such as Microsoft, Amazon, Intel, Meta, and Google. Additionally, he explores novel data compression methods for data systems and implements adaptive compression selection in both conventional and resource-constrained databases and machine learning systems. Furthermore, he is engaged in high-dimensional data analysis, with a particular emphasis on time series applications.

News

Mar 25': Our demo paper on a chat interface for declarative AI frameworks, “PalimpChat: Declarative and Interactive AI Analytics,” was accepted by SIGMOD 2025.
Feb 25': Our paper on scientific discovery “Variable Extraction for Model Recovery in Scientific Literature” was accepted by AISD @ NAACL 2025.
Feb 25': Our paper on open columnar formats evaluation focusing on embedding and feature store workloads“Data Formats in Analytical DBMSs: Performance Trade-offs and Future Directions” was accepted by the VLDB Journal special issue “Best of VLDB.”
Jan 25': Our paper on LLM scheduling “ Don't Stop Me Now: Embedding Based Scheduling for LLMs” has been accepted to ICLR 2025.

Publications

Professional Contributions

Professional Services

Program Committee of SIGMOD'25, VLDB'24-26, KDD'2023, SMDB'2023
Reviewer of The International Journal on Very Large Data Bases (VLDBJ)
SIGMOD Availability & Reproducibility Committee 2022, 2023

Awards

GUIDEAI@SIGMOD Best Paper Award 2024
VLDB Best Paper Runner-Up 2023
University Unrestricted Fellowship 2019

Teaching Assistant

CMSC 23500 Introduction to Databases 2021, 2020, 2019 Winter, University of Chicago
CMSC 15200 Introduction to Computer Science II 2017 Spring, University of Chicago
CMSC 23010 Parallel Computing 2017 Winter, University of Chicago
CMSC 23000 Operating Systems 2016 Autumn, University of Chicago
712061H Mining Massive Data 2015 Spring, University of Chinese Academy of Sciences
711003H Computer Network 2014 Autumn, University of Chinese Academy of Sciences

	Chunwei Liu Postdoctoral Associate at MIT CSAIL 32 Vassar St, Cambridge, MA 02139 Office: Room 32-G930 MIT Data Systems and AI Lab (DSAIL) MIT Data Systems Group (DSG) Email: chunwei@csail.mit.edu Profiles: GoogleScholar, LinkedIn, Twitter ORCID, ResearchGate, CV

Short Bio Chunwei Liu is a Postdoctoral Associate at MIT CSAIL working with Michael Cafarella. He got his Ph.D. from Department of Computer Science at the University of Chicago, where he worked in ChiData group advised by Aaron Elmore. His research interests span compound AI systems, database systems, cloud/edge computing, and database benchmarking. He focuses on optimizing data systems for both conventional data analytics and emerging AI-powered pipelines. Chunwei develops privacy-preserving workload generation techniques for evaluating cloud database systems, collaborating with major cloud vendors such as Microsoft, Amazon, Intel, Meta, and Google. Additionally, he explores novel data compression methods for data systems and implements adaptive compression selection in both conventional and resource-constrained databases and machine learning systems. Furthermore, he is engaged in high-dimensional data analysis, with a particular emphasis on time series applications. News Mar 25': Our demo paper on a chat interface for declarative AI frameworks, “PalimpChat: Declarative and Interactive AI Analytics,” was accepted by SIGMOD 2025. Feb 25': Our paper on scientific discovery “Variable Extraction for Model Recovery in Scientific Literature” was accepted by AISD @ NAACL 2025. Feb 25': Our paper on open columnar formats evaluation focusing on embedding and feature store workloads“Data Formats in Analytical DBMSs: Performance Trade-offs and Future Directions” was accepted by the VLDB Journal special issue “Best of VLDB.” Jan 25': Our paper on LLM scheduling “ Don't Stop Me Now: Embedding Based Scheduling for LLMs” has been accepted to ICLR 2025. Publications Professional Contributions Professional Services Program Committee of SIGMOD'25, VLDB'24-26, KDD'2023, SMDB'2023 Reviewer of The International Journal on Very Large Data Bases (VLDBJ) SIGMOD Availability & Reproducibility Committee 2022, 2023 Awards GUIDEAI@SIGMOD Best Paper Award 2024 VLDB Best Paper Runner-Up 2023 University Unrestricted Fellowship 2019 Teaching Assistant CMSC 23500 Introduction to Databases 2021, 2020, 2019 Winter, University of Chicago CMSC 15200 Introduction to Computer Science II 2017 Spring, University of Chicago CMSC 23010 Parallel Computing 2017 Winter, University of Chicago CMSC 23000 Operating Systems 2016 Autumn, University of Chicago 712061H Mining Massive Data 2015 Spring, University of Chinese Academy of Sciences 711003H Computer Network 2014 Autumn, University of Chinese Academy of Sciences