What is our primary use case?
CAST AI serves as our primary solution for Kubernetes cost optimization and automated node management in our EKS cluster.
One example of how we use CAST AI for cost optimization and node management is in our production EKS environment where workload fluctuates throughout the day. Before CAST AI, we manually sized node groups and often provisioned over-provisioned resources. CAST AI automatically provisions the most cost-effective instances and continuously right-sizes the cluster based on workload demand. This significantly reduced unused capacity while maintaining application performance.
We use CAST AI daily to monitor cluster efficiency, optimize resource allocation, and reduce the operational effort required to manage Kubernetes infrastructure.
What is most valuable?
The best features CAST AI offers, in my experience, are automated Kubernetes cost optimization, intelligent auto-scaling, spot instance management, workload right-sizing recommendation, and cluster visibility.
Automated node provisioning and optimization has made the biggest difference for us. It reduced the need for manual intervention and helped ensure we are always running the most cost-efficient infrastructure.
CAST AI has positively impacted our organization by reducing cloud costs, improving resource utilization, and allowing our engineering team to spend less time managing infrastructure and more time on platform improvements.
What needs improvement?
I would like to see more granular reporting, deeper cost allocation insights, and additional customization options for optimization policies in CAST AI.
Overall the platform is strong, and most improvements needed would be around reporting and advanced governance capabilities for larger organizations.
Regarding CAST AI's AI capabilities, the governance and security controls are solid. It provides sufficient visibility into cluster changes and optimization actions, although more advanced policy would be beneficial.
Enhanced forecasting capabilities and more detailed workload-level cost analytics would be useful improvements for CAST AI that I have not mentioned yet.
For how long have I used the solution?
I have been using CAST AI for more than a year.
What do I think about the stability of the solution?
CAST AI has proven to be stable and reliable in production environments.
What do I think about the scalability of the solution?
CAST AI's scalability is very good; it scales effectively with cluster growth and increasing workload complexity.
How are customer service and support?
Our experience with customer support has been positive. Response times are reasonable and the team is knowledgeable.
Which solution did I use previously and why did I switch?
Before CAST AI, we relied mainly on Kubernetes auto-scaling and manual monitoring processes.
Before choosing CAST AI, we looked at AWS native optimization tools and a few Kubernetes cost optimization platforms before selecting CAST AI due to its automation capabilities.
What was our ROI?
I have seen a return on investment, and the ROI was visible within a few months through cloud cost reduction alone. Additionally, our team spends less time manually managing Kubernetes infrastructure.
What's my experience with pricing, setup cost, and licensing?
My experience with pricing, setup cost, and licensing was that the setup process was straightforward, pricing was reasonable considering the cost savings achieved, and licensing was easy to understand.
What other advice do I have?
Since adopting CAST AI, we achieved approximately 30-40% reduction in Kubernetes infrastructure costs. We also reduced manual cluster management activities significantly, especially around node scaling and capacity planning.
The best features CAST AI provides are automated Kubernetes cost optimization, intelligent auto-scaling, spot instance management, cluster visibility and analytics, and workload right-sizing recommendations.
Regarding the accuracy and reliability of CAST AI's AI capabilities, recommendations are generally accurate and reliable. We always validate major changes, but in most cases, the optimization suggestions are practical and effective.
My advice to others looking into using CAST AI is to start with a non-production cluster to understand the optimization recommendations, establish baseline cost metrics, and then granularly expand adoption across environments.
Overall, CAST AI has been a valuable addition to our Kubernetes platform operations. It helped us reduce cloud spending while simplifying cluster management, and I would recommend it to organizations looking to optimize Kubernetes cost at scale. I rated CAST AI as an eight out of ten.
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?
Amazon Web Services (AWS)