Experience

SyntheticGestalt

• United Kingdom
• Biotechnology
• 1 - 100 Employee

• Research Engineer

  • Sep 2022 - Present

  Developing cloud-based machine learning systems for drug discovery. - Deep learning-based hit generation algorithm development - ADMET prediction algorithm development - Quick model prototyping with Amazon Sagemaker (with ECR and S3) - Collaborative model development using Docker and devcontainer in VSCode Developing cloud-based machine learning systems for drug discovery. - Deep learning-based hit generation algorithm development - ADMET prediction algorithm development - Quick model prototyping with Amazon Sagemaker (with ECR and S3) - Collaborative model development using Docker and devcontainer in VSCode



National Institute of Advanced Industrial Science and Technology (AIST)

• Japan
• Research Services
• 700 & Above Employee

• Visiting Ph.D. student

  • Apr 2020 - Mar 2023

  As a visiting PhD student, I worked on developing deep learning-based efficient design methods for biological sequences (specifically proteins) at the Artificial Intelligence Research Center. Supervised by Dr. Yutaka Saito. As a visiting PhD student, I worked on developing deep learning-based efficient design methods for biological sequences (specifically proteins) at the Artificial Intelligence Research Center. Supervised by Dr. Yutaka Saito.



PKSHA Technology

• Japan
• IT Services and IT Consulting
• 1 - 100 Employee

• Senior algorithm engineer

  • Jan 2017 - Sep 2022

  In PKSHA, I have a track record of developing various machine learning algorithms. My clients were the top-tier companies in automobile, public security, telecommunication, pharma/healthcare, and many other industries. Mostly used PyTorch for deep learning but sometimes TensorFlow v1/v2, OpenCV, DGL for graph-structured data, and PyData stack for analytics. Selected past project topics: - Object/person detection algorithm for surveillance - Interpretable human pose/action recognition algorithm in the wild - Implementation of unsupervised visual representation learning algorithm - Serving ML models using Docker and gRPC/REST API in production environments - Automating (pipelining) data preprocessing and ML model training/evaluation with Apache Airflow - Creating an easy-to-use annotation tool for human pose recognition (with Vue.js and Flask) - ... and many more projects that cannot be disclosed Show less



M3

• Fundraising
• 1 - 100 Employee

• Data Analyst

  • Jan 2016 - Jan 2017

  With Python and its data science stack (numpy, scipy, pandas, scikit-learn, SQLAlchemy), I * developed machine learning engines for better ads display on web sites or for recommender systems * developed some systems for monitoring/reporting KPIs over several web services of our own * performed statistical tests on marketing campaigns With Python and its data science stack (numpy, scipy, pandas, scikit-learn, SQLAlchemy), I * developed machine learning engines for better ads display on web sites or for recommender systems * developed some systems for monitoring/reporting KPIs over several web services of our own * performed statistical tests on marketing campaigns



スローガン株式会社

• Japan
• Staffing and Recruiting
• 1 - 100 Employee

• Director

  • Nov 2012 - Jan 2016

  Provided HR services for early-stage companies. Specifically, - Consulting for graduate recruitment - Planning, writing, and editing corporate branding materials, as the editor in chief - Web application development using Python for in-house data management etc. Provided HR services for early-stage companies. Specifically, - Consulting for graduate recruitment - Planning, writing, and editing corporate branding materials, as the editor in chief - Web application development using Python for in-house data management etc.