What is our primary use case?
My main use case for Pulumi Enterprise is that it is an infrastructure as code tool. We're slowly importing legacy infrastructure and creating all new infrastructure with Pulumi.
A quick specific example of how I'm using Pulumi Enterprise in my infrastructure is that we provision all AWS resources and generally all cloud resources with Pulumi, including SSO, EKS cluster, databases, and everything. Everything is managed by Pulumi.
Something unique about my use case with Pulumi Enterprise is how it manages secrets. It's quite convenient as it is able to encrypt secrets right in the repository with a KMS key. Additionally, AI works quite well with Pulumi. It helps generate code and it works well for the development team. They prefer it much more than Terraform. It is more aligned with what they see daily, making it a more developer-friendly toolset.
What is most valuable?
The best features Pulumi Enterprise offers for my team is a combination of a high-level infrastructure as code tool and all the capability of a general programming language. We use Pulumi runtime for TypeScript and Pulumi with Python.
Using TypeScript or Python with Pulumi makes things easier for my team because it is fast and reliable, and the team is able to use a language that they are familiar with. This is important when the infrastructure or platform team collaborates closely with the back-end team as they don't need a translation layer between Terraform and what they work with. TypeScript really helps there. Additionally, Pulumi has been linked to all major programming languages, making it definitely helpful.
Pulumi Enterprise has impacted my organization positively as it is a modern infrastructure as code tool that helps us manage resources efficiently and reduce delivery time.
What needs improvement?
Regarding improvements for Pulumi Enterprise, it would be beneficial to move a Pulumi-generated plan from an experimental feature to general availability. It works, but I don't feel confident when I have to enable an experimental flag in a production environment.
For how long have I used the solution?
I have been using Pulumi Enterprise for approximately three years.
What do I think about the stability of the solution?
Pulumi Enterprise is stable enough.
What do I think about the scalability of the solution?
Pulumi Enterprise's scalability is limited as it is not particularly scalable. You should keep the project relatively small to avoid simultaneous Pulumi runs.
How are customer service and support?
Regarding customer support, we use the open-source version, so I have never communicated with customer support.
Which solution did I use previously and why did I switch?
I had experience with various infrastructure as code tools with my previous employers, and we also used Terraform in some projects. There was no particular reason why we switched to Pulumi; we started using it from the beginning in that infrastructure. We simply tried it, and it works well.
What was our ROI?
I have seen a return on investment using Pulumi as we need fewer employee resources, and we spend much less time on routine tasks. We probably saved two engineer times with Pulumi.
Which other solutions did I evaluate?
Before choosing Pulumi Enterprise, I evaluated other options, including Terraform.
What other advice do I have?
Pulumi has reduced delivery time by transitioning from a lengthy process to provision a new environment to a repeatable process. You do it once and then do it again and again.
On a scale of one to ten, I would rate Pulumi Enterprise an eight.
I say eight because this annoying experimental feature keeps it from being a ten. We really want to have a deterministic plan, review the deterministic plan before we apply it in production, and now it is possible only when we enable it as an experimental feature, which I don't feel comfortable doing in a production environment. The second consideration is that sometimes readability could be better, and I have experienced some issues with Pulumi state, though I don't recall exact details.
Regarding Pulumi Enterprise's AI capabilities, I think it provides a pretty good output; however, I haven't used it thoroughly, so I'm not familiar with this part of the product.
I actually had some experience with Pulumi Enterprise AI capability, and I found that its output is pretty inaccurate compared to a general AI model. I stopped using it, and they probably have improved it significantly in the last couple of years.
We don't purchase Pulumi Enterprise; we use the open-source version.
The advice I would give to others looking into using Pulumi Enterprise is that it is one of many infrastructure as code tools. Try it and see if it works for your particular use case. The main advantage is that the team is able to write code in a general programming language, not a domain-specific language like Terraform. On a scale of one to ten, I rate Pulumi Enterprise an eight.
Which deployment model are you using for this solution?
Public Cloud
If public cloud, private cloud, or hybrid cloud, which cloud provider do you use?