End to End Possibility for Dev-Ops Implementation with Best Practices

Amrit

Write now Dataiku possess a lot of unique abilities to develop scalable ML / Deep Learning / GenAI algorithms. In addition to that Dataiku has facilitated collaborative development using flow-zones etc. Even there are a lot of Data quality checks and metrics to facilitate operational efficiency and drift detection.

But all in all, Dataiku's DevOps practices does not seem to be following the best practices as per today's DevOps standard. It is understandable as including visual recipes is not a easy ft. Right now Dataiku documentation suggests using Remote Git but handling Visual Recipes in Remote Git is not ideal. Apart from that, writing automation for testing, governance scripts again do not follow the best practices.

Developing a flow with end to end best practices will go a long way into Dataiku's industrial level adoption.

Turribeach

It's not really clear what you are asking here. While in principle I agree that Dataiku is harder to integrate with a DevOps pattern it is not imposible. The level of integration that your DevOps solution can have will depend on how much you want to do, the Dataiku features you use and the technologies you want to integrate with. For that reason is hard to have a single DevOps guide. There are however several CI/CD examples in the knowledge site that should review.