What is our primary use case?
We use Pulumi Enterprise with the Python runtime to provision our Azure platform, essentially our managed identities, CosmoDB containers, their data planes and role assignments, Key Vault access, event hub RBAC, and that sort of layer. It's split into per-service stacks at the moment that each own their own resources inside shared infrastructure, so it means one team can essentially add what they need without touching another's.
How has it helped my organization?
Pulumi Enterprise has allowed us to be much more confident in the code changes that we are actually deploying and releasing, and just as an example, I mentioned earlier, we're now unit testing all of our Pulumi Enterprise infrastructure as part of the CI pipeline. So we know quite confidently that the change is actually going to be successful before it's actually been deployed, which has been hugely beneficial for us.
It's definitely saved time and accuracy in the development space because we're now spending less time testing, debugging code, and worrying about manual infrastructure. We now do all of the DevOps ourselves, so definitely saved time there. I wouldn't really want to give a specific number on it here, but it has been very beneficial.
What is most valuable?
Another good point for us is that we're on Pulumi Cloud as an organization, so state is managed rather than us babysitting a blob storage backend, and every stack has a browsable update history. For example, I just ran a prod change this morning that was update 10 on that stack, and it gave me a shareable URL for the preview diff and the apply. The feature we lean on most is ESC, their environments and config layer, so we get one shared environment that supplies the tenant subscription ID and cluster coordinates to every stack that imports it, meaning there's nothing environment specific, essentially, which is a huge benefit for us.
The main feature we use, to be honest, is state management.
What needs improvement?
I think debugging can sometimes be less intuitive; for example, because you're executing actual code to produce infrastructure, there's a mental gap between the code I'm writing, the Pulumi Enterprise engine, the desired state that we're after once the code actually executes, the provider, for example, in this case Azure, and the cloud resource. When something doesn't behave as expected, there's another layer of abstraction to understand. You're not just debugging your code; you're debugging how Pulumi Enterprise interprets that code and interacts with the provider as well. So it can be an extra mental layer to get your head around.
The quality of the experience with Pulumi Enterprise definitely depends somewhat on the provider and the resource that you're working with. For us with Azure, we've found it to be very useful, but I could definitely see that if there isn't the same sort of level of quality with other providers, then we would probably have a more negative experience.
In engineering, there's always something that can be improved even if the consumers don't know what they need, and I'm sure there is definitely room for improvement.
For how long have I used the solution?
I've been using Pulumi Enterprise for around nine or 10 months.
What do I think about the stability of the solution?
Pulumi Enterprise is stable.
What do I think about the scalability of the solution?
We find Pulumi Enterprise very scalable.
How are customer service and support?
I've actually never needed to reach out to customer support, so I'm not sure, but I would imagine it's quite good because it feels like quite a modern system with modern support.
Which solution did I use previously and why did I switch?
We used Terraform, and we switched because we just found the state management quite cumbersome. New people that were trying to use Terraform got quite confused by that, and also it wasn't really testable, and we found that our complicated business logic wasn't really suitable for Terraform.
How was the initial setup?
There were no issues with the initial setup that I could think of.
What about the implementation team?
I wasn't too involved with that part of the process.
What was our ROI?
The ROI was quite good, actually.
Which other solutions did I evaluate?
We did evaluate other options before choosing Pulumi Enterprise, but I can't remember now; as I say, it's been nine or 10 months, and they weren't good enough to remember.
What other advice do I have?
I would rate Pulumi Enterprise an eight out of ten.
I've been using Pulumi Enterprise with Claude as a plugin for a while, and to be honest, I found the governance and security very good with Claude Code. I haven't really been concerned about it at all, and it's been very good at what it does.
It's been very accurate to me; previously to this, I used Copilot with Terraform, and I found that to be very inaccurate. So coming from that to Pulumi Enterprise and Claude Code has been wonderful.
I'd definitely recommend installing the Claude plugin and just getting a general overview on how Pulumi Enterprise can help improve your infrastructure.
Which deployment model are you using for this solution?
Private Cloud
If public cloud, private cloud, or hybrid cloud, which cloud provider do you use?
Microsoft Azure