Jaehyun Ha

Department of Artificial Intelligence, POSTECH, South Korea

Portrait of Jaehyun Ha

Research Interests

Building AI-native database systems for unified analytics and inference on multi-modal data

Graph Databases
Architecting a foundational query framework based on a unified graph representation of multi-modal data
Semantic Operator
Embedding Large Language Models (LLMs) into the database to enable context-aware querying of unstructured data
Query Optimization
Designing proxy models and novel cardinality estimation methods to optimize the execution of computationally expensive semantic operators

Publications

  1. 2027

    CADENZA: Compiling Natural-Language Intent into Task-Specific Operator DAGs for Semantic Query Processing.Jaehyun Ha, Yongjoo Park, Wook-Shin Han.SIGMOD 2027.

  2. 2026

    CADENZA in Action: Breaking the Monolith with Intent-Dependent Plan Spaces for Semantic Queries.Jaehyun Ha, Yongjoo Park, Wook-Shin Han.VLDB 2026 (Demo).

  3. 2026

    TurboLynx: Schemaless Graph Engine Strikes Back for General-Purpose Analytics.Taesung Lee, Jaehyun Ha, Byungchul Tak, Wook-Shin Han.VLDB 2026.

  4. 2026

    TurboLynx in Action: A Schemaless Graph Engine for General-Purpose Analytics.Taesung Lee, Jaehyun Ha, Byungchul Tak, Wook-Shin Han.VLDB 2026 (Demo).

  5. 2024

    DoppelGanger++: Towards Fast Dependency Graph Generation for Database Replay.Wonseok Lee*, Jaehyun Ha*, Wook-Shin Han, Changgyoo Park, Myunggon Park, Juhyeng Han, Juchang Lee. (* equal contribution)SIGMOD 2024.

  6. 2024

    DoppelGanger++ in Action: A Database Replay System with Fast Dependency Graph Generation.Wonseok Lee, Jaehyun Ha, Wook-Shin Han, Changgyoo Park, Myunggon Park, Juhyeng Han.VLDB 2024 (Demo).

  7. 2023

    Guaranteeing the O(AGM/OUT) runtime for uniform sampling and size estimation over joins.Kyoungmin Kim, Jaehyun Ha, George Fletcher, Wook-Shin Han.PODS 2023.

Professional Experience

  1. Google Systems Research Group

    Student Researcher · Sunnyvale, CA, USA

    • Researching hardware-software co-optimization to accelerate semantic queries
    • Hosted by Yeounoh Chung; collaborated with Helena Caminal and Fatma Özcan
  2. University of Illinois Urbana-Champaign (UIUC)

    Visiting Scholar · Urbana–Champaign, IL, USA

    • Conducted research on semantic operator optimization, resulting in a paper accepted at SIGMOD 2027
    • Joint research project under the supervision of Prof. Yongjoo Park
  3. Digital Platform team, SK hynix Inc.

    Software Engineer Intern · Icheon, Korea

    • Developed real-time video comment overlay system for SK hynix’s internal video streaming systems
    • Won silver prize (2nd) in internship contest

Education

  1. POSTECH, Graduate School of Artificial Intelligence

    M.S./Ph.D. Integrated Program · Data Systems Lab

    Advisor: Prof. Wook-Shin Han · Expected: Feb. 2028

  2. POSTECH, Dept. of Computer Science and Engineering

    B.S. in Computer Science and Engineering · GPA: 4.06 / 4.3

    Graduated with the highest GPA in CS Dept.

  3. Ulsan Science High School

    Graduated early