ML Platform
Last Updated :
Summary
Mercari built a platform to promote machine learning to make our infrastructure and computer resources operate efficiently. Our ML engineers carry out experiments and data processing efficiently and safely to give us the flexibility to quickly apply results to production.
Purpose
We develop and improve Mercari's ML platform. This platform includes the ML infrastructure and orchestration systems provided with all the functionality for our ML engineers to run experiments using our data.
Description
Machine learning encompasses many kinds of components as well as learning models. To reproduce experiments, it is essential to have a stable environment which allows you to develop and execute models. The ML platform is designed so that ML engineers can concentrate on developing models and constructing a continuous pipeline. Automating most orchestration simplifies the development process.
ML engineers can develop models regardless of infrastructure or resource requirements. They can also test a new model or easily create a notebook for carrying out a data survey. Even from an ideation phase, our engineers can swiftly transplant a project to an end-to-end ML model pipeline having full functionality.