Philipp Koehn
Professor at Johns Hopkins University
Cockeysville, Maryland, United States
Summary
🤩
Rockstar🎓
Top SchoolPhilipp Koehn is a professor and seasoned machine translation researcher with 14+ years of experience bridging academia and industry, currently affiliated with Johns Hopkins and serving as Chief Scientist at Omniscien Technologies. He has held senior research roles at Facebook AI and long-standing professorships at the University of Edinburgh, combining deep theoretical expertise with practical system-building. His contributions to the widely used Moses MT toolkit show a focus on robust, production-ready scoring and parallelization improvements that prevent edge-case failures. Trained with a PhD in Computer Science from USC and advanced degrees from European institutions, he brings a rare blend of rigorous research, open-source stewardship, and product-minded engineering to large-scale language technologies.
14 years of coding experience
10 years of employment as a software developer
PhD, Computer Science, PhD, Computer Science at University of Southern CaliforniaDiplom, Computer Science, Diplom, Computer Science at FAU Erlangen-NürnbergAlbert Schweitzer Gymnasium, ErlangenMSc, Computer Science, MSc, Computer Science at University of Tennessee, Knoxville
German, English, Latin, Portuguese
Github Skills (10)
algorithm10
numerical-optimization10
code-optimization10
algorithms10
machine-translation10
perl10
optimisation10
optimization10
numerical-analysis9
test-automation7
Programming languages (5)
Github contributions (5)
moses-smt/mosesdecoder
Dec 2011 - Sep 2016
Moses, the machine translation system
Role in this project:
Back-end Developer
Back-end DeveloperContributions:260 commits, 27 pushes in 4 years 9 months
Contributions summary:Philipp contributed to the Moses machine translation system by updating the scoring functions within the `mteval-v13a.pl` file. These updates involved modifying the scoring functions to prevent division-by-zero errors when a system segment is empty. Additional changes include preparations for parallel processing, and improvements to file paths and default settings. This user appears to be focused on maintaining and improving the core performance and robustness of the scoring mechanisms within the system.
i18nmachine-translationtranslationlocalizationmoses
casmacat/moses-cat-server
Jun 2013 - May 2016
Contributions:16 commits, 2 pushes in 2 years 11 months
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.