This question, that I posed on the mlops.community Slack, was recently highlighted in their monthly newsletter:
The end of a recent podcast (Coffee Session #94 minute 41:48) touches on this a bit, but nothing is clearly defined yet. As a consultant with multiple clients across varying companies and ML technologies, I find myself managing projects that are at different stages of the MLOps lifecycle.
I’m leaning towards putting together a dashboard (in Notion, or Trello, for example) around Google Cloud’s "Practitioner’s Guide to MLOps" - to help me keep track of where each project is in the stage of the MLOps lifecycle. I’d love to be able to trace the project back to customer requirements, and later on when in production, to use ML monitoring tools (to track for drift, etc).
Do other MLOps practitioners track their projects at a similar level? Like a S/W product manager might?