JL.

Junfeng Liu · NTU Singapore

Junfeng Liu

PhD candidate in database systems

I am a PhD candidate at Nanyang Technological University, advised by Prof. Siqiang Luo. My research focuses on workload-aware storage structures, LSM-tree optimization, and graph storage.

I am part of the Graph Analysis and Data System Lab (GADSL), where we work on scalable data systems, advanced data structures, and graph algorithms.

Selected publications Email GitHub

01

Research

A

Adaptive LSM-tree systems

Re-architecting compaction, indexes, caches, and level structures around changing workloads, memory budgets, and data modalities.

MooseArceKVAdCache

B

Graph storage

Graph-aware layouts for disk-resident and in-memory systems that balance update throughput, scan locality, and space.

AsterRadixGraph

C

Learned & multimodal storage

Studying learned indexes and adaptive layouts for heterogeneous data and evolving query modes in agentic systems.

Learned indexesAI for systems

02

Selected publications

SIGMOD2026

RadixGraph: A Fast, Space-Optimized Data Structure for Dynamic Graph Storage

Haoxuan Xie, Junfeng Liu, Siqiang Luo, and Kai Wang

Fast updates and compact indexing for dynamic in-memory graphs.

SIGMOD2025

Aster: Enhancing LSM-structures for Scalable Graph Database

Dingheng Mo, Junfeng Liu, Fan Wang, and Siqiang Luo

A polymorphic LSM layout for billion-scale graph workloads.

03

Background

Education

  1. Nanyang Technological UniversityPhD, Database Systems
  2. The University of ManchesterMaster of Science
  3. Sun Yat-sen UniversityBachelor of Science

Industry

PingCAPDatabase Engineer · 2022 — 2023

Worked on TiDB data movement and cloud infrastructure across TiDB Lightning, TiFlow, TiCDC, TiDB Operator, and Kubernetes.

Tools I work with

C++ · Go · Python · SQL · Linux · RocksDB · Docker · Kubernetes · performance profiling

04

Contact

I am always happy to discuss database systems, storage engines, and possible research collaborations.

junfeng001@e.ntu.edu.sg