Switching from amateur projects to enterprise scale machine learning.
People transition into machine learning from diverse backgrounds and most often they are software developers; Nevertheless, everyone begins with personal or curricular projects. In the industry, however, the projects are at a larger scale and bring additional complexities into their development and maintenance. This is where a fully managed ML service or ML Ops (inspired from DevOps) comes in. It shifts the process to the cloud and offers greater flexibility, reliability and transparency to the team. Through this article, I draw parallels between software development and typical data science projects.