JUN HYUK CHANG
He is a fifth-year Ph.D. student in Computer Science at the University of Chicago, advised by Aaron J. Elmore and collaborating closely with Goetz Graefe. Before his Ph.D., he earned his M.S. and B.S. from KAIST and spent about five years in industry at SK hynix and a database systems company.
- Implemented batch ANN (Approximate Nearest Neighbor) vector search in F1 Query and deployed it to production.
- Benchmarked on Gemini pre-training data, cut CPU time by up to 50x over kNN, and reached 94% recall with 25x fewer workers.
- Optimized Bolt, ByteDance's Velox-based OLAP query execution engine, on Apache Spark and Presto.
- Implemented a zero-copy serializer/deserializer using Apache Arrow to accelerate data spilling.
- Designed and developed platform firmware (FIL and FTL) for NAND flash memory semiconductors.
- On leave for Ph.D. studies since Sep. 2022.
- Led a team of 10 engineers owning the database's active storage layer.
- Built the cloud storage manager, including sharding and re-balancing.
- Built real-time interactive advertising and voting systems for live streams.
- Designed and verified NAND read/write functionality through simulation and testing with Verilog and HSPICE.
Enhancing Transaction Processing through Indirection Skipping 
VLDB 2025: Proceedings of the Very Large Data Base Endowment Inc, Volume 18, Issue 11, Pages 4104–4116
Resource-Adaptive Query Execution with Paged Memory Management
CIDR 2025: Conference on Innovative Data Systems Research
Towards Resource-adaptive Query Execution in Cloud Native Databases
CIDR 2024: Conference on Innovative Data Systems Research
Amygdala resting-state functional connectivity alterations in patients with chronic insomnia disorder: correlation with electroencephalography beta power during sleep
Sleep 2023: Sleep, Volume 46, Issue 10, October 2023, zsad205
Enjoy the Silence: Noise Control with Smartphones
ICCCN 2017: IEEE International Conference on Computer Communications and Networks
AIFI: Asynchronous Interaction Feedback Interface for Remote Presentation Preparation
KSC 2017: Korean Institute of Information Scientists and Engineers (KIISE), Korea Software Congress