About
Hi, I'm Sam! I'm a fifth-year PhD student in Computer Science at Carnegie Mellon University (CMU), co-advised by Prof. Andy Pavlo, Prof. Todd Mowry, and Prof. Jignesh Patel.
My research explores how compiler optimization techniques can improve the performance of user-defined functions (UDFs) invoked by analytical SQL queries.
Before starting grad school, I worked as a compiler engineer at Apple, where I helped ship Metal-accelerated rendering in Blender 3.1. Before that, I did my undergrad at The University of Sydney, Australia.
Research Impact
During my PhD, I discovered weaknesses in existing subquery unnesting algorithms, which motivated improved approaches in Umbra, DuckDB, and Microsoft Fabric.
Open Source Contributions
During an internship at Columnar, I built the DuckDB Arrow Database Connectivity (ADBC) extension, which lets DuckDB connect to any of the 30+ database systems with ADBC drivers. Earlier in my PhD, I also implemented nested LATERAL joins in DuckDB with Mayank Baranwal and Arham Chopra, which shipped in DuckDB v0.9.
Selected Publications
The Key to Effective UDF
Optimization: Before Inlining, First Perform Outlining
Runner-Up Best Paper Award 🏆
SIGMOD Research Highlight Award 🏆
Samuel Arch, Yuchen Liu, Jignesh M. Patel, Todd C. Mowry, Andrew Pavlo.
VLDB 2025
Dear User-Defined Functions,
Inlining isn’t working out so great for us. Let’s try batching to make our relationship work.
Sincerely, SQL
Kai Franz, Samuel Arch, Denis Hirn, Torsten Grust, Todd C. Mowry, Andrew Pavlo.
CIDR 2024
Experience
PhD Student, CMU (2022 - Present).
Software Engineering Intern, Columnar (2025 - Present).
High-Performance Query Processing Intern, Amazon Redshift (2024).
Compiler Engineer, Apple (2021).
Teaching
15-418: Parallel Computer Architecture and Programming (Spring '25).
15-745: Optimizing Compilers for Modern Architectures (Fall '23).
Awards
Graduate Fellowship in the School of Computer Science, CMU (2025).
The University Medal, The University of Sydney (2020).