Domino has been around since 2013, which in this field makes it ancient, and it has survived by picking a problem that does not go out of fashion. The problem is not building models. Plenty of tools build models. The problem is proving to somebody with statutory authority that you know exactly which model produced which decision, trained on which data, approved by whom, and that you can reproduce it two years later when the question is asked.
Pharmaceutical companies and financial institutions have this problem acutely and it is not a nice-to-have. Domino's governance centre, model registry and audit trail exist because of that requirement and they are noticeably better thought through than the governance features bolted onto general purpose ML platforms. Credit where it is due. The platform side is solid and unexciting in the way infrastructure should be.
On-demand compute, environments that are reproducible rather than aspirational, support for Python and R and the usual frameworks without forcing you into a proprietary notation, app hubs for putting models in front of business users, and deployment across on-premise, hybrid and multicloud. The multicloud and on-premise options are not a checkbox here, they are the reason a bank can use it. The investor list including Sequoia, NVIDIA and Snowflake tells you where the company is positioned. Now the part relevant to a site about learning.
Domino offers courses and certifications through its learning portal, and the documentation is thorough. My honest assessment is that this training is not portable. It teaches you to operate Domino, and Domino is not a skill you carry between jobs the way Python or Kubernetes or general MLOps practice is. Without a platform to work on, the courses are abstract, and you cannot casually get a platform because there is no free tier and no public pricing.
So the calculus is simple. If you have joined a company that runs Domino, work through the training and get certified, because you will be more useful faster and the certification signals commitment internally. If you are building a career and choosing what to learn next, this is not it. Learn the underlying disciplines, model governance as a concept, experiment tracking, reproducible environments, deployment patterns, and those will transfer to whichever platform your next employer bought.
The absence of public pricing deserves a comment because it tells you something. Enterprise software that will not publish a number is priced per customer based on what that customer can bear, and in this category that means six figures for a meaningful deployment. That is not a criticism of the product, which genuinely does things that justify enterprise pricing, but it does mean this is a decision made by a CTO rather than a practitioner, and if you are a practitioner your input is limited to whether you can work with it once it arrives. You can.
It is a good platform. It is just not a thing you learn on your own time. Three point five for learning value, materially higher as a platform if governance is your binding constraint.