Cast AI - EKS fully automated cost optimization and monitoring

My main use case was to migrate development and staging environments to Kubernetes using CAST AI, and with them, we created these environments. We moved them from virtual machines to Kubernetes clusters.

CAST AI helped with that migration by creating a Kubernetes cluster designed to automatically manage infrastructure components. This allowed us to boost the project I work on inside Zazmic. This allowed us to focus on application deployment rather than complex infrastructure provisioning. The team also gave direct feedback to us when we were implementing. They developed specific scripts to automate the creation. This enabled the team to easily spin up or tear down environments on demand as needed for testing and validations.

CAST AI positively impacted our organization by leading the migration of a very high number of pre-production environments from VMs to Kubernetes. We had these specific sandboxes for each developer, and there was a high volume of environments. This helped us make it easier to administer and manage each environment for each developer. It really gave us an opportunity to manage our systems better. Regarding cost and savings, CAST AI's optimization software helped us move away from a mess of machines in our environments to really make progress on that matter, to manage our costs better, and have a more maintained environment, and use our resources better.

CAST AI can be improved by managing the role-based access control better. I think it would be better for CAST AI to improve how access permissions are handled for developers. That would be a beneficial refinement on that matter. Managing node management better is important. We had a recurring issue with our nodes where they were becoming not ready and required manual deletion to be reclaimed by the cluster. Making this automation better would be great. The cost optimization is great, but improving the automation implementation of cost-saving measures, such as weekend scaling, would be something useful as well.

We began to work with CAST AI on April 2021.

CAST AI is stable.

The customer support is great. The customer support and the team effort to help us put the cluster up or maintain it is great. It is really useful.

Before choosing CAST AI, we were evaluating putting up the Kubernetes clusters by ourselves, but we chose CAST AI to help us on that matter first.

We had monthly savings of 2,500 to 3,000 from our bills on cloud.

My experience with pricing, setup cost, and licensing was that the proposal they shared with us included an ongoing monthly cost. There was a one-time setup cost as well, but all that we spent with CAST AI was okay because we have monthly savings that surpassed what we spent with the tool.

Automation and infrastructure optimization features helped my team by creating the cluster to manage infrastructure components. This allowed the booster project to focus on application deployment rather than complex infrastructure provisioning. These features made it easier for the administration of the creation of the cluster and administration of it. We did not have to spend so much time going through the cloud itself to provide this for us.

The best features CAST AI offers include automated infrastructure management, infrastructure optimization, and tooling for automation. These are things that stand out. The possibility to use solutions integrated with it is particularly helpful.

The migration to Kubernetes helped reduce cloud costs by enabling the use of preemptible instances and optimizing cluster infrastructure.

Regarding CAST AI's AI capabilities, I think its governance and security are well-managed. We did not have issues with how it is handled and how we give people access to the systems or to CAST AI itself.

CAST AI's capabilities are really accurate and reliable.

I would rate this product a 9 overall.

Public Cloud

Amazon Web Services (AWS)

What do you like best about the product?

What I love most about Cast AI is the automated, real-time right-sizing. Before using it, our engineering team spent countless hours manually looking at Grafana dashboards, trying to guess which AWS EC2 instances to use, and wrestling with Kubernetes cluster sizing.

Cast AI completely takes that off our plate. It analyzes our actual resource demand in real time and automatically swaps out inefficient nodes for optimized, cost-effective instances without any downtime. The fact that it manages Spot Instances so smoothly—automatically moving workloads to On-Demand nodes if a Spot instance gets interrupted—has given us massive savings on our cloud bill without sacrificing application stability. It’s truly a "set-it-and-forget-it" optimization tool.

What do you dislike about the product?

What I love most is how it optimizes Kubernetes on autopilot. It watches our traffic in real time and automatically swaps out nodes for cheaper, perfectly sized ones without any downtime. It even manages Spot instances seamlessly, slashing our cloud bill without the stress of crashes.

The biggest downside is the pricing—they charge based on your total cluster size, not just the money they save you, so the bill gets expensive fast as you scale. Plus, the automation can be a bit of a black box during sudden traffic spikes, and fixing it means digging through some really messy documentation.

What problems is the product solving and how is that benefiting you?

The biggest problem Cast AI solves is cloud waste from over-provisioning. Our team used to buy way more AWS compute than we needed just to avoid crashes, but Cast AI automatically scales our nodes up and down to match actual demand in real time.

It also takes the risk out of using cheap Spot instances. It predicts when AWS is about to reclaim an instance and moves our workloads to a fresh node before anything drops. For me, that means instantly slashing our cloud bill by thousands of dollars without the stress of managing infrastructure or dealing with late-night alerts.

What do you like best about the product?

I use CAST AI to optimize Kubernetes clusters by automatically reducing cloud costs, right-sizing workloads, and improving application performance. It's especially valuable for production workloads where balancing performance, reliability, and infrastructure costs is important. What I like most about CAST AI is how much of the Kubernetes optimization process it automates. The platform continuously analyzes workloads and makes intelligent scaling and right-sizing decisions. I also appreciate the clear cost visibility, support for Spot instances, and actionable recommendations that help reduce cloud spending without compromising application reliability or performance. Additionally, it was very easy to set up.

What do you dislike about the product?

Overall, CAST AI has been very effective, but there are a few areas that could be improved. The platform has a lot of powerful features, so the initial learning curve can be a bit steep for teams that are new to Kubernetes optimization. I'd also like to see more guided onboarding, additional best-practice templates for common cluster configurations, and more detailed cost forecasting before optimization changes are applied. Expanding the reporting dashboard with deeper historical trends, workload-level insights, and clearer explanations for automated recommendations would also make it easier to understand and trust optimization decisions.

What problems is the product solving and how is that benefiting you?

I use CAST AI to optimize Kubernetes clusters, reducing cloud costs and improving performance. It automates resource management, scaling, and optimizes node sizing, saving me time on infrastructure tasks and letting me focus on building and deploying applications.

What do you like best about the product?

I like CAST AI for its automated optimization, which has made a significant impact on managing multiple client clusters efficiently. Its ability to analyze resource usage and optimize node provisioning helps minimize infrastructure costs and reduces cloud waste. The automation feature enables rightsizing of workloads, maintaining efficiency at scale, and identifying underutilized compute resources. The initial setup process was smooth, making the transition from manual Kubernetes optimization seamless.

What do you dislike about the product?

I think more granular client level reporting, better forecasting capabilities, and enhanced workload recommendations can be improved in the platform. Also, the initial setup process was smooth, but if you are running with multi-client with multi-environment, it will be a learning curve for new users.

What problems is the product solving and how is that benefiting you?

I use CAST AI to optimize Kubernetes workloads, minimize infrastructure cost, and reduce cloud waste. It rightsizes workloads, automates node optimization, and saves effort in managing multiple client clusters, improving efficiency at scale.

My primary use case for CAST AI is Kubernetes cost efficiency, and since I handle the cloud system as well, CAST AI has been instrumental in helping me. We mainly use it to automate processes in AWS EKS clusters.

For my recent use case with CAST AI for Kubernetes cost efficiency in AWS EKS clusters, we are building a sourcing platform in our production environment, where we have deployed an EKS environment filled with workloads. Before CAST AI, we mainly sized node groups, which often led to overpricing. CAST AI automatically provided us with a clearer vision of our clusters based on the use case we are addressing. This is our main use case and example.

The main and biggest feature of CAST AI is that with its help, I am able to automate Kubernetes, and because of that, cost efficiency and cost optimization are significantly better than before, along with workload balancing and intelligent automations. These are the main features.

CAST AI has positively impacted our organization in all three areas, where it has reduced the cost of our cloud infrastructure, the manpower that we were previously applying to optimize anything, and utilization is very much improved as we spend less time managing infrastructure.

In the last Q2 result, because of using CAST AI, we have reduced our manpower, money, and cost by 20 to 30%, which indicates substantial funding reduction.

I would like to see CAST AI improved with deeper and more intelligent answers and solutions, along with additional optimization and customization options. The customization option in particular could be enhanced to help further.

Overall, the platform is very strong, and most improvements could include advanced customization, advanced reporting, and documentation on a large scale.

We have been using CAST AI for more than one year.

My advice to others looking into using CAST AI is that if you are new to this and do not know much, you can use CAST AI to learn things and to gain hands-on experience on production level applications.

Before concluding, I would like to say that if you are new to this, CAST AI is very efficient before making a decision, and it is also very good from a cost point of view, saving considerable resources overall. I would rate this review a 9 out of 10.

Private Cloud

Amazon Web Services (AWS)