How Recommendations are Made?

After you link your AWS accounts to our platform, an intricate process begins. Through the AWS EC2 APIs, we analyze your account data, focusing on your on-demand EC2 instances.

The data is fed to our predictive models, which are specifically trained to evaluate your usage patterns and predict your future needs. Based on this finetuning, we identify saving opportunities for your infrastructure to optimize your costs mainly by transitioning to reserved instances.

Why don’t recommendations match my instances?

This happens if your instance is too large, we split it into smaller reserved instances or if it’s too small we combine multiple on-demand instances with one large reserved instance.

Example: You might be running an m5.xlarge instance and receive a recommendation to get TWO m5.large reserved instances.

We do this to better optimize the reserved capacity that you have, so if you decided to move some of your instances to another region, another account under your organization, or if you decided to downgrade your on-demand instance to a smaller size, it will be more efficient to handle these changes with the right reserved instance size.

Does that impact my on-demand instance?

We don’t interfere with your running on-demand instances in any way. The process of switching to reserved instances based on our recommendations involves securing a reserved instance contract that applies only to billing. This means you will get cost reductions without any changes to your instances performance, configurations, or availability.