MLOps Engineer

The MLOps Engineer uses machine learning on JMDC's healthcare data to create new value. The role includes data-analysis consulting for insurance companies and healthcare businesses, building predictive models, providing services, and improving the value of services operated by business units and group companies.

The engineer is expected to advance projects independently and autonomously, from problem definition through social implementation. The position provides experience in developing and operating services using data science and machine learning, analyzing a large healthcare database, and applying statistical and machine-learning skills. Healthcare and medical-domain knowledge can also be developed through the work.

Required qualifications include logical thinking, experience working in a team according to a Git branching strategy such as Git Flow or GitHub Flow, and experience designing, developing, and operating externally facing API services while considering non-functional requirements. Experience designing, developing, and operating systems using PaaS or FaaS is required. The role also requires experience applying infrastructure as code and CI/CD based on DevOps principles, including log collection and performance monitoring.

Preferred experience includes delivering and continuously improving production services using MLOps, implementing machine-learning models or methods proposed in recent papers, placing in data-analysis competitions such as Kaggle, developing healthcare-focused data-science or machine-learning models, and having specialist biostatistics or clinical knowledge for medical research. The posting also seeks strong oral and written communication, an interest in new things, and a commitment to self-development.

The listed technologies include Python and SQL; Amazon SageMaker, AWS Fargate on ECS, AWS Lambda, AWS Athena, Amazon Redshift Serverless, Amazon DynamoDB and Vertica; Terraform; GitHub Actions and AWS CodeBuild; and tools including Gemini Advanced, NotebookLM, ChatGPT Plus, GitHub Copilot, Slack, Backlog, Confluence, Google Workspace and GitHub Enterprise. Windows notebook and computing desktop equipment or a MacBook Pro is provided.

This is a full-time permanent position with a monthly salary system. Salary is undisclosed and determined according to company rules based on experience and ability. There is a three-month probationary period, potentially extended by up to three additional months, with the same conditions. Overtime is paid separately, and expected overtime is approximately 10 to 20 hours. The work location is the Tokyo office in Minato, Tokyo. Working hours use a flextime system with core hours from 11:00 to 15:00; standard working time is eight hours per day with a 60-minute break. Benefits and systems stated include social insurance, commuting allowance up to 150,000 JPY per month, a defined-contribution pension, leave programs, an employee stock ownership plan and health-related support programs.