Kyubyong Park
Chief Executive Officer at TUNiB
Gyeonggi, South Korea
Summary
🤩
Rockstar🎓
Top SchoolKyubyong Park is a computational linguist and CEO with nine years of hands-on experience applying deep learning to NLP and speech, now leading TUNiB after research leadership at Kakao Brain. He combines rigorous academic training (M.A. in Linguistics, SNU undergrad) with prolific open-source engineering—implementing TensorFlow versions of Transformer, Tacotron, DC-TTS, and practical G2P systems on his GitHub. His work spans end-to-end TTS, grapheme-to-phoneme conversion, multilingual word vectors, and lightweight speaker-adaptive voice cloning, reflecting a rare blend of lexicography expertise and production-ready ML engineering. He’s also an experienced language educator and author of more than ten Korean textbooks for learners, which informs his applied approach to linguistic problems. Based in Gyeonggi (Seoul), he runs open deep learning projects full-time and often ships reproducible research and training code that others can build on. Colleagues describe him as a practitioner-researcher who moves quickly from linguistic insight to deployable model.
10 years of coding experience
13 years of employment as a software developer
Bachelor's degree, Korean Language and Literature, Bachelor's degree, Korean Language and Literature at Seoul National UniversityMaster of Arts (M.A.), Linguistics, 3.97, Master of Arts (M.A.), Linguistics, 3.97 at University of Hawaii at Manoa
Korean, English, Japanese, Indonesian, Spanish, German, Chinese
Stackoverflow
1reputation
0reached
0answers
0questions
Github Skills (24)
python10
architecture10
attention-mechanism10
data-manipulation10
machine-learning10
text-to-speech10
rnn-model10
speech-synthesis10
n10
numpy10
tensorflow10
speech10
natural-language-processing10
neural-network10
transformer10
Programming languages (5)
Github contributions (5)
Kyubyong/transformer
Jun 2017 - Feb 2019
A TensorFlow Implementation of the Transformer: Attention Is All You Need
Role in this project:
ML Engineer
ML EngineerContributions:22 commits, 6 PRs, 27 pushes in 1 year 8 months
Contributions summary:Kyubyong primarily contributed to the implementation of a Transformer model using TensorFlow. Their work involved defining core components such as embedding layers, multi-head attention mechanisms, and feedforward networks. They also integrated positional encoding and implemented label smoothing for training. The user's commits focused on building the model architecture and configuring the training process.
tensorflowattention-mechanismtranslationattention-is-all-you-needtransformer
Kyubyong/dc_tts
Nov 2017 - Apr 2018
A TensorFlow Implementation of DC-TTS: yet another text-to-speech model
Role in this project:ML Engineer
ML EngineerContributions:39 commits, 1 PR, 14 pushes in 4 months
Contributions summary:Kyubyong primarily contributes to the implementation and modification of a deep learning model for text-to-speech (TTS), as evidenced by the code diffs. Their work involves defining and structuring neural network components like text encoders, audio encoders, and decoders, as well as integrating attention mechanisms. The commits reveal the refinement of network architectures, integration of dropout layers, and modification of the loss functions within the training process.
tensorflowtext-to-speechspeechspeech-to-texttts
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.