What is our primary use case?
My main use case for Infracost is checking the cost of adding any new resources using infrastructure as code, and this application runs some pipelines on GitHub and shows the cost of this particular operational action that we are going to do.
I can give you a quick specific example of a recent time I used Infracost in this way: we use Azure, and there are different infrastructures that need to be provisioned to keep our environment running smoothly. We do a lot of sizing and scaling of our infrastructure, and all of this is done through infrastructure as code, especially using Terraform or another IaC tool. Whenever we make a change, before we actually apply it, Infracost runs pipelines and lets us know what the added cost or removed cost of this particular operation is. If we're adding infrastructure or scaling up or scaling down a server, Infracost shows us the exact cost of this particular action. We usually scale up when we expect traffic to increase and scale down when the traffic decreases.
The primary use case from Infracost's side is understanding our provisioning cost and then looking at reducing the cost or optimizing it.
How has it helped my organization?
Infracost has positively impacted my organization by enabling us to reduce our cloud bill without impacting performance, achieving at least a 20% reduction in the amount we spend on cloud. It also helps us shift left by allowing us to understand the costs before we actually provision it.
I achieved that 20% cost reduction primarily by addressing overprovisioning, and also because Azure comes with different types of server types and app service types. The AI capabilities of Infracost help by recommending specific SKUs, which contributes to reducing costs.
What is most valuable?
The best features Infracost offers are the product itself and its easy integration in an infrastructure as code platform directly, especially with Terraform, and it understands Azure, since we are on Azure. It's easy to integrate, easy to understand, and easy to use out of the box.
The integration is easy to configure, and now it comes with some AI enablement that automatically tells us if something is overprovisioned or underprovisioned.
Regarding Infracost's AI capabilities, I think it comes with a lot of governance checks and is quite secure. I don't see a lot of issues here.
Infracost's AI capabilities are very well-trained and accurate. I have not seen any hallucinations or deviations.
What needs improvement?
Infracost is actually a very good product; however, the improvement part could be on the pricing side because the non-free plans are a bit expensive and don't necessarily offset the savings.
Feature-wise, I don't have a lot of complaints about this product; it's quite amazing with a good concept. I do think that with AI catching up, tools such as ChatGPT or Codex will be able to do what Infracost does, so they need to come up with something to differentiate their product.
I choose that assessment because of the pricing part; if the pricing was much better, since the starter pack is $250 a month, I'm not sure we are always able to save that much every month, making it an expense instead of a saving.
The only improvements needed for Infracost are on the pricing part.
For how long have I used the solution?
I have been using Infracost for a couple of years.
What do I think about the stability of the solution?
Infracost is very stable, extremely stable.
What do I think about the scalability of the solution?
Infracost's scalability is very good; it all depends on the price plan, as we can run it for various amounts such as 1,000 runs or 10,000 runs.
How are customer service and support?
I haven't had to interact with customer support yet.
Which solution did I use previously and why did I switch?
I did not previously use a different solution.
How was the initial setup?
The setup process was actually very easy for my team; there were no challenges or surprises. It's integrated on Azure, and we see the report on the GitHub Actions that we run. Whenever a pipeline runs on GitHub Action, it sends us the report for a server, showing how much additional cost, such as two dollars or five dollars, it adds to the monthly bill or reduces the monthly bill.
What was our ROI?
I have seen a return on investment; as I mentioned, we have achieved a 20% saving in our infrastructure costs. Our ability to scale up and down our infrastructure before making any changes means we know the pricing and predicted price very well, understanding how much it's adding to the cost. The only concern remains the starter price of $250 per month, which we aren't sure we can always save.
What's my experience with pricing, setup cost, and licensing?
The pricing needs a lot of improvement; otherwise, the setup cost, licensing, and everything else is very easy, very smooth, and very developer-friendly. They know their audience very well.
Which other solutions did I evaluate?
Before choosing Infracost, I didn't evaluate other options; we just heard about it and decided to go with it.
What other advice do I have?
To others looking into using Infracost, I would say it is a really good product. Infracost's free plan comes with a sufficient number of runs, 1,000 runs, which is very generous. It has good integrations, so I recommend going ahead with it if you're looking to save on your cloud bill and shift left. I would rate this product an 8 out of 10.
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?
Microsoft Azure