Рет қаралды 66
Presented by: Conrado Miranda, CTO at Verta
There’s a fundamental misunderstanding that plagues many companies on their ML journey. Oftentimes, businesses try to adapt tried-and-true DevOps processes used to write traditional software to execute the unfamiliar task of Machine Learning operationalization (aka MLOps). But a copy/paste approach between DevOps and MLOps rarely works. In this talk, we’ll draw a parallel between developing a library and developing a model - and show why Machine Learning requires a new approach.