Teaching + Supervision

During my PhD study, I have been enjoying teaching and organizing teaching activities, such as leading the TA team and managing exercises and examinations. It helps me gain valuable experience in management and commnication.

Here is the list of courses I have taught so far.

• Data Management Systems: Head TA & Invited Speaker (Fall 2024), TA (Fall 2025)
• Hardware Acceleration for Data Processing Seminar: Invited Speaker (Fall 2024)
• Data Modeling and Databases: TA (Spring 2024, 2025, 2026)
• Information Systems for Engineers: TA (Fall 2023)

I am also very fortunate to work with talented students on various projects.

Master’s Thesis

• Sepehr Pourghannad (now as a performance engineer at IMC Trading): Integrated Eiger, a high-performance GPU-based query engine, into Maximus, a modular database working with heterogeneous hardware.
• Dominic Steiner (now as a software engineer at Optiver). Profiled the NCCL library for the distributed database workloads and proposed optimizations that greatly improve the distributed operator’s performance in Eiger.

Bachelor’s Thesis

• Michael Simon: Characterizing semantic operations in databases and co-designing databases and LLM inference.

Semester Projects

• Ana Paladi: Implementing in-memory compressed data storage and processing for GPU-accelerated databases.
• Elvan Konukseven: Accelerating database operators on Hopper architecture.
• Phillip Trummer: Characterized agentic text-to-SQL workflow and exploring databases and LLM agents co-design.
• Cong Vu: Studied unified memory and memory spilling for out-of-memory data processing on the GPU platform.
• Diana Micloiu: Ported Eiger from CUDA to ROCm, profiling performance differences, and performing optimizations.
• Marios Angelis (Incoming intern at IBM research center): Implemented high-performance GPU-based database operators, such as left-join, semi-join, and anti-join, in Eiger.