$ cat cv_en.md
Yuki Kashiwada/ 柏田祐樹
ML Engineer · Medical AI Researcher · CTO / Board Member
contact@ai-driven.jp · github.com/javasparrows · researchmap.jp/yukikashiwada
Last updated: 2026-07-14
##Experience
Feb 2025 – Present
HR Service Company— AI Engineer (Contract)
AI development lead for multiple generative-AI product features. Built LLM-based chat and document-revision features and shipped them into a service used by 100k+ users. (Next.js / FastAPI / GCP / Vertex AI / Gemini, team of 5)
Dec 2024 – Present
BINARISTA, Inc.— Board Member / PdM & AI Engineer
Product management, development, and operation of an enterprise sales-support AI-agent SaaS. Built the product from scratch and acquired new customers. (Next.js / FastAPI / LangChain / LangSmith / Gemini API / PostgreSQL, team of 5)
Oct 2024 – May 2025
SherLOCK, Inc.— CTO
Led AI-security consulting and product development. Conducted LLM security assessments and won a joint-research project with a research institute and a PoC with a major enterprise. (Next.js / FastAPI / OpenAI API, team of 4)
2023 – Present
RIKEN Center for Advanced Intelligence Project (AIP), Statistical Genetics Team— Research Part-time Worker I
Engaged in medical AI research in the Statistical Genetics Team.
Jun 2023 – Mar 2024
Entertainment Company— AI Development Lead (Contract)
Developed person-detection and image-processing AI, improving detection accuracy from 90% to 96%. Owned planning through implementation. (PyTorch / TensorFlow / GCE, team of 8)
Dec 2022 – Nov 2024
Climate-tech SaaS Company— AI Engineer (Contract)
Built an NLP-based invoice-PDF recognition AI, improving multi-field extraction accuracy from 40% to 80%. (PyTorch / spaCy / GCP, team of 5)
Jan 2021 – Dec 2022
NTT DATA Corporation— Senior Data Scientist
AI development lead on a smart pig-farming project. Designed and built an end-to-end AI pipeline, improving accuracy from 65% to 90% and reaching on-site deployment. Also launched a new data-science team. (PyTorch / TensorFlow / Databricks / React / Flask, team of 10)
Aug 2019 – Dec 2020
NABLAS Inc.— AI Engineer
Built anomaly-detection systems for manufacturing image and time-series data, achieving 85%+ accuracy in automotive-part defect detection. (PyTorch / TensorFlow / scikit-learn / tslearn, team of 5–7)
##Education
Oct 2022 – Present
Tohoku University Graduate School of Medicine— Ph.D. Course, Medical Science (expected Sep 2026)
Medical AI research: breast cancer detection in mammography, EEG analysis, and LLM applications to electronic health records.
Apr 2017 – Mar 2019
The University of Tokyo, Graduate School of Science— M.Sc., Astronomy
Apr 2009 – Mar 2014
Waseda University— B.Sc., Earth Science, School of Education
##Publications & Datasets
2025
K. Konomatsu, Y. Kashiwada, T. Kubota, et al. — Epilepsy & Behavior, 172:110713
Deep learning on density spectral arrays of single-channel ictal EEG to differentiate epileptic from functional/dissociative seizures (AUC 0.941).
2025
Y. Kashiwada, E. Takaya, M. Hiroya, et al. — PeerJ Computer Science, 11:e3252
Lead author. Proposed a model integrating ViT and CNN to capture multi-scale morphological features in mammography for breast cancer detection.
2025
Y. Kashiwada, E. Takaya, M. Hiroya, et al. — The Cancer Imaging Archive (TCIA)
Lead author. An annotated mammography dataset for developing and validating breast cancer detection AI, publicly available on TCIA.
##Presentations
2025
退院サマリへの大規模言語モデル適用による症例自動抽出
柏田祐樹 ほか — 日本メディカルAI学会 学術集会
Automated case extraction from discharge summaries using large language models (paper in preparation).
2024
Vision Transformerによるマンモグラフィ乳癌識別モデルの開発
柏田祐樹 ほか — 学会発表
The paper based on this work was published in PeerJ Computer Science (MIME-ViT).
2018
Effects of galactic structures on solar motion measurements
柏田祐樹 ほか — 学会発表
Presentation on the effects of galactic structures on solar motion measurements (astronomy).
##Awards
May 2025
MedTech Angels Season 4— Special Award (BD Fund and RD Support)
Jan 2025
MiTOHOKU 2024 (Regional MITOU Program)— Grand Prize
Sep 2024
Japan Epilepsy Society— Excellent Poster Award
Dec 2023
Tohoku University Business Idea Contest 2023 (personalized epilepsy medicine)— Grand Prize & EY Award
##Patents
2024
Information processing method for brain disease assessment (JP App. 2024-35856)— Patent application
##Certifications
Apr 2025
Information Security Management Examination (SG)— IPA
Nov 2024
JDLA Deep Learning for ENGINEER (E-shikaku)— JDLA
Jun 2024
Generative AI Test— JDLA
Oct 2022
Kaggle Expert (4 silver medals)— Kaggle
Mar 2022
JDLA Deep Learning for GENERAL (G-kentei)— JDLA
Nov 2021
Data Scientist Certification (Literacy Level)— Data Scientist Society
Apr 2021
Duolingo English Test 115 (≈ TOEFL 90)— Duolingo
Mar 2021
TOEIC 840— ETS
##Other Activities
2025
OWASP— Supervising translator, OWASP Top 10 for LLM Applications 2025 (Japanese edition)
2023 – Present
Tohoku University School of Medicine— Co-founding member, personalized epilepsy medicine team "perple"
R&D on personalized epilepsy medicine using genetic data. Awarded 60M JPY over 3 years by the Michinoku Gap Fund (STEP2).
2024 – Present
Japanese Society for Medical AI— Member
##Skills
Generative AI / LLM
AI agent development / LangChain / LangSmith / OpenAI API / Gemini / Vertex AI / RAG / LLM security (OWASP LLM Top 10 translation)
Machine Learning
PyTorch / TensorFlow / scikit-learn / Computer vision / NLP / Time-series analysis / Accuracy improvement (40%→80%, 90%→96%, etc.) / Kaggle Expert
Medical AI
Mammography image analysis / EEG analysis / LLM on electronic health records / Joint research with physicians, peer-reviewed publications
Software Engineering
Python / TypeScript / Next.js / React / FastAPI / PostgreSQL / MySQL / AWS / GCP / Docker / Mobile app development
Management
Product management / Leading teams of 5–10 / Full lifecycle from requirements to operation / Client-facing planning and proposals / CTO / board experience / Business communication in English