avatar

Saehan Jo

Applied Scientist, AWS
PhD, Cornell University
sj683 (at) cornell.edu


I graduated in the summer of 2025 and joined the Learned Systems Group at AWS!

About Me

I am an Applied Scientist in the Learned Systems Group at AWS. I did my PhD in Computer Science at Cornell University, where I was fortunate to be advised by Prof. Immanuel Trummer.

Research Interests

I am particularly interested in building efficient data systems involving LLMs and ML models. At Cornell, I developed an LLM framework that automatically selects minimum-cost LLMs while ensuring sufficient result quality. I also built an approximate query processing system for multi-modal data on top of an RDBMS, ML models, and LLMs.

Work Experience

During my PhD, I interned with the Learned Systems Group at Amazon, where I worked with Kapil Vaidya, Murali Narayanaswamy, and Prof. Tim Kraska on LLM pipelines for Text2SQL (Amazon Q generative SQL). I also interned with the Data Systems Group at Microsoft Research, collaborating with Tarique Siddiqui, Wentao Wu, and Chi Wang on query workload compression for index tuning.

Publications

  1. SpareLLM: Automatically Selecting Task-Specific Minimum-Cost Large Language Models under Equivalence Constraint
    Saehan Jo, Immanuel Trummer
    Proceedings of the ACM on Management of Data (SIGMOD 2025).
    PDF
  2. ThalamusDB: Approximate Query Processing on Multi-Modal Data
    Saehan Jo, Immanuel Trummer
    Proceedings of the ACM on Management of Data (SIGMOD 2024).
    PDF Code
  3. Demonstration of ThalamusDB: Answering Complex SQL Queries with Natural Language Predicates on Multi-Modal Data (Demo)
    Saehan Jo, Immanuel Trummer
    Companion of the 2023 International Conference on Management of Data (SIGMOD 2023).
    PDF Code
  4. ISUM: Efficiently Compressing Large and Complex Workloads for Scalable Index Tuning
    Tarique Siddiqui, Saehan Jo, Wentao Wu, Chi Wang, Vivek Narasayya, Surajit Chaudhuri
    Proceedings of the ACM on Management of Data (SIGMOD 2022).
    PDF
  5. SkinnerDB: Regret-bounded Query Evaluation via Reinforcement Learning
    Immanuel Trummer, Junxiong Wang, Ziyun Wei, Deepak Maram, Samuel Moseley, Saehan Jo, Joseph Antonakakis, Ankush Rayabhari
    ACM Transactions on Database Systems (TODS 2021).
    PDF 'Best of SIGMOD' Edition
  6. BitGourmet: Deterministic Approximation via Optimized Bit Selection
    Saehan Jo, Immanuel Trummer
    Conference on Innovative Data Systems Research (CIDR 2020).
    PDF
  7. Demonstration of BitGourmet: Data Analysis via Deterministic Approximation (Demo)
    Saehan Jo, Immanuel Trummer
    Proceedings of the ACM on Management of Data (SIGMOD 2020).
    PDF
  8. Demonstration of ScroogeDB: Getting More Bang For the Buck with Deterministic Approximation in the Cloud (Demo)
    Saehan Jo, Jialing Pei, Immanuel Trummer
    Proceedings of the VLDB Endowment (VLDB 2020).
    PDF
  9. Mining an "Anti-Knowledge Base" from Wikipedia Updates with Applications to Fact Checking and Beyond
    Georgios Karagiannis, Immanuel Trummer, Saehan Jo, Shubham Khandelwal, Xuezhi Wang, Cong Yu
    Proceedings of the VLDB Endowment (VLDB 2019).
    PDF
  10. AggChecker: A Fact-Checking System for Text Summaries of Relational Data Sets (Demo)
    Saehan Jo, Immanuel Trummer, Weicheng Yu, Xuezhi Wang, Cong Yu, Daniel Liu, Niyati Mehta
    Proceedings of the VLDB Endowment (VLDB 2019).
    PDF
  11. Verifying Text Summaries of Relational Data Sets
    Saehan Jo, Immanuel Trummer, Weicheng Yu, Xuezhi Wang, Cong Yu, Daniel Liu, Niyati Mehta
    Proceedings of the ACM on Management of Data (SIGMOD 2019).
    PDF
  12. SkinnerDB: Regret-Bounded Query Evaluation via Reinforcement Learning
    Immanuel Trummer, Junxiong Wang, Deepak Maram, Samuel Moseley, Saehan Jo, Joseph Antonakakis
    Proceedings of the ACM on Management of Data (SIGMOD 2019).
    PDF
  13. SkinnerDB: Regret-Bounded Query Evaluation via Reinforcement Learning (Demo)
    Immanuel Trummer, Samuel Moseley, Deepak Maram, Saehan Jo, Joseph Antonakakis
    Proceedings of the VLDB Endowment (VLDB 2018).
    PDF
  14. Fast and Scalable Distributed Loopy Belief Propagation on Real-World Graphs
    Saehan Jo, Jaemin Yoo, U Kang
    Proceedings of the ACM International Conference on Web Search and Data Mining (WSDM 2018).
    PDF Code
  15. Supervised Belief Propagation: Scalable Supervised Inference on Attributed Networks
    Jaemin Yoo, Saehan Jo, U Kang
    IEEE International Conference on Data Mining (ICDM 2017).
    PDF Code

Powered by Jekyll and Minimal Light theme.