Luke Hsiao
Software Engineer
Maritime Stealth Startup

I am a husband, father, and engineer that enjoys building well-designed, ethical systems that serve the greater good. Currently, I am a member of technical staff at an early-stage stealth startup in the maritime industry where I'm helping develop technology I believe will provide broad societal benefits.

Previously, I helped build the foundations of Alation's AI platform; I joined Alation as part of its acquisition of Numbers Station, where I designed and implemented cloud infrastructure and backend services from the company's first day through its acquisition. Prior to Numbers Station, I was on the Network Infrastructure Group at Google and worked on TCP in Linux.

I completed my Ph.D. in electrical engineering at Stanford University, advised by Phil Levis and Keith Winstein and with the collaboration of Chris Ré. My dissertation focused on making electronic component information more accessible by using machine learning to build knowledge bases directly from PDF datasheets. While in grad school, I also researched low-latency, foveated video compression for future VR systems, built drivers at NVIDIA, and implemented optimizations for Linux networking at Google.

Before Stanford, I graduated summa cum laude from Brigham Young University with a B.S. in computer engineering and worked mainly on embedded systems and FPGA reliability with Mike Wirthlin.

Research

Select publications, in reverse chronological order.

Towards Retina-Quality VR Video Streaming: 15ms Could Save You 80% of Your Bandwidth
L. Hsiao, B. Krajancich, P. Levis, G. Wetzstein, and K. Winstein
ACM SIGCOMM Computer Communication Review, Jan. 2022

code press release

Creating Hardware Component Knowledge Bases with Training Data Generation and Multi-task Learning
L. Hsiao, S. Wu, N. Chiang, C. Ré, and P. Levis
ACM Transactions on Embedded Computing Systems (TECS), Sept. 2020

code data

Automating the Generation of Hardware Component Knowledge Bases
L. Hsiao, S. Wu, N. Chiang, C. Ré, and P. Levis
Languages, Compilers, and Tools for Embedded Systems (LCTES), Jun. 2019

code data slides poster

Fonduer: Knowledge Base Construction from Richly Formatted Data
S. Wu, L. Hsiao, X. Cheng, B. Hancock, T. Rekatsinas, P. Levis, and C. Ré
ACM Conference on Management of Data (SIGMOD), May 2018

code slides poster talk

Misc. projects