Machine Learning Engineer (Image Processing)

This role develops machine-learning models and image-processing algorithms for solutions using satellite imagery. Responsibilities include preprocessing satellite images, improving image quality, and building data pipelines for service delivery.

The work focuses on earth-observation data obtained through remote sensing. This includes imagery from visible light, infrared, and SAR wavelengths, together with complex metadata such as geographic information and capture times. The role may explore ways to combine data from different satellites and sensors with terrestrial IoT data and apply satellite imagery to different regions and business domains.

The position does not require prior knowledge of space engineering; that knowledge can be learned after joining. The product development team belongs to the Data Solutions Department of the Solution Ecosystem Business Division and currently consists of approximately three to four members around age 30.

The main programming language is Python, with Rust and TypeScript also used. Development environments are macOS or Linux, including WSL2. Communication uses GitHub, Slack, and Zoom. Code is managed with GitHub, task management uses GitHub Projects V2, GitHub Issues, and Redmine, version control uses Git, CI/CD uses GitHub Actions, and infrastructure as code uses Terraform.

The work location is the Ariake office in Koto-ku, Tokyo. Remote work is available, while the average schedule involves approximately 3.5 days per week in the office, adjustable to individual circumstances. The role can involve work across satellite development, operation, data utilization, constellation systems, and software applications for global projects.