MLOps Platform for Chip Manufacturers
**Problem** The client aimed to create a user-friendly web platform for their MVP, building upon existing machine learning code. This platform would enable users to upload datasets, register microchips, and conduct scalable end-to-end training experiments in the cloud. Prior to the project, model training was partially manual, and there was a lack of a cohesive connection between the platform and the machine learning experiments. Our collaborative objective was to integrate all components of the platform, introduce necessary new features, and implement authentication and authorization systems.
**Solution** A team of four engineers from Brainstorm Developers focused on various aspects of the project, including the frontend (using ReactJS), backend (NodeJS), bridge code between the backend and Ray Cluster (Python), and infrastructure (AWS). Among the completed features are: the entire platform is hosted on AWS, utilizing Amazon Aurora MySQL for the database, EC2 instances for model training with Ray for machine learning experiments, S3 for storage, and DataSync for automated data transfers.
**Results** The Brainstorm Developers play a vital role in product design, decision-making, and planning. Development is carried out using the Scrum agile project management framework, with tools like Slack, Jira, and GitHub. As of this writing, the collaboration is ongoing, and the outlook is promising.