Takahiro Omi
Research Manager at ストックマーク株式会社
Chiyoda, Japan
Summary
👤
Senior🎓
Top SchoolTakahiro Omi is a research leader specializing in natural language processing and large language models, currently VP of Research at Stockmark where he leads development of business-domain LLMs such as Stockmark-100b and Japan-focused 13B models. He combines deep academic grounding—a PhD from Kyoto University and an eight-year tenure as a specially appointed associate professor at the University of Tokyo—with product-focused machine learning engineering and industry-academia collaborations. His work spans information extraction from news, ML platform engineering, and generative AI research, and he co-authored practical textbooks on BERT and Transformers. Omi has a track record of publishing in top venues like NeurIPS and translating research into operational systems for seismic analysis and financial domains. Colleagues note his ability to bridge rigorous statistical time-series modeling with scalable NLP productization. He is based in Chiyoda, Tokyo and brings a rare combination of academic depth, technical authorship, and leadership in commercializing large-scale language models.
9 years of coding experience
1 year of employment as a software developer
博士（理学）, 理学研究科, 博士（理学）, 理学研究科 at 京都大学
Github Skills (15)
korean-nlp10
similarity10
temporal9
nlp9
transformers8
wikipedia8
natural-language-processing8
bert7
transformer7
pytorch6
deep-learning6
neural-network4
python2
simulation2
inference1
Programming languages (2)
Github contributions (5)
omitakahiro/NeuralNetworkPointProcess
May 2019 - Jan 2021
code for "Fully Neural Network based Model for General Temporal Point Processes"
Contributions:16 commits, 1 PR, 14 pushes in 1 year 8 months
pytorchmodel-basedpoint-processesdeep-learningtemporal
stockmarkteam/bert-book
Feb 2021 - Apr 2022
「BERTによる自然言語処理入門: Transformersを使った実践プログラミング」サポートページ
Contributions:49 commits, 14 PRs, 41 pushes in 1 year 2 months
nlptransformersbertsimilaritytransformer
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.