Matthias Sperber
Siri Machine Translation R&D Scientist
Aachen, North Rhine-Westphalia, Germany
Summary
🤩
Rockstar🎓
Top SchoolMatthias Sperber is a Siri Machine Translation R&D Scientist at Apple with nine years of experience at the intersection of research and production NLP. He brings a strong academic foundation from KIT and hands-on international research exposure at CMU and NAIST, translating cutting-edge sequence modeling into robust systems. His open-source contributions to DyNet—implementing and optimizing LSTM internals, multi-dimensional operations, and dropout—demonstrate deep expertise in low-level neural network engineering often invisible in higher-level model work. Based in Aachen, he excels at bridging research prototypes and production-grade ML for real-world speech and translation products.
9 years of coding experience
5 years of employment as a software developer
Master of Science (MS), Computer Science, Master of Science (MS), Computer Science at Karlsruhe Institute of TechnologyBachelor of Science (BS), Computer Science, Bachelor of Science (BS), Computer Science at Technische Universität Kaiserslautern
German, English, Japanese
Github Skills (14)
neural-network10
pytorch10
machine-learning10
lstm10
c-language10
deep-learning10
tensorflow10
recurrent-neural-networks10
cprogramming-language10
cuda9
algorithms8
data-structures8
algorithm8
data-structure8
Programming languages (2)
Github contributions (5)
clab/dynet
May 2017 - May 2018
DyNet: The Dynamic Neural Network Toolkit
Role in this project:
ML Engineer
ML EngineerContributions:45 commits, 18 PRs, 10 pushes in 1 year
Contributions summary:Matthias made significant contributions to the DyNet toolkit by implementing and improving LSTM nodes, crucial for recurrent neural networks. They focused on developing the forward and backward passes for these nodes, including the LSTM gates, optimizing performance, and addressing bugs. The user's work also involved integrating multi-dimensional operations and refining dropout mechanisms within the LSTM framework, demonstrating expertise in deep learning implementation.
dynetdynamic-neural-networkdeep-learningneural-networksmachine-learning
msperber/xnmt
Mar 2018 - Feb 2020
eXtensible Neural Machine Translation
Contributions:4 PRs, 35 pushes, 7 branches in 1 year 11 months
machine-translationtranslationtorchnmtneural
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.