Bridging the Gap: MLOps Transformations in Streamlining Model Deployment

Maria recounts the outdated approach of zipping code and emailing it to a separate DevOps team, emphasizing the inefficiencies of the previous method. Basak explains the major change implemented, which involved breaking down barriers between data scientists and DevOps engineers. By taking ownership of deploying models within the same team, they eliminated the need for error-prone handovers between teams. This shift to MLOps facilitated streamlined and collaborative model deployment processes.

MLOps Coffee Sessions #159 with Maria Vechtomova, Lead ML engineer, …

breaking change code data data scientists deployment devops devops engineers engineers gap major mlops model deployment ownership scientists team