Quote 1:
“Currently we have a system that resembles half-MLOps but lacks data versioning tied to model versioning and configurations, and makes it difficult to deploy and keep track [of AI projects].”
Majid, Data Scientist @ Siemens
Quote 2:
"As we have an open source first policy we are forced to go with tools like MLflow. But I really don’t like their 'opinionated' approach how practitioners should log artifacts. I like the idea of using existing best practices in the MLOps space."
Quote 1: “Currently we have a system that resembles half-MLOps but lacks data versioning tied to model versioning and configurations, and makes it difficult to deploy and keep track [of AI projects].”
Quote 2: "As we have an open source first policy we are forced to go with tools like MLflow. But I really don’t like their 'opinionated' approach how practitioners should log artifacts. I like the idea of using existing best practices in the MLOps space."