Koyeb has announced the general availability of its autoscaling feature, which aims to provide a global, serverless experience for deployments by automatically adjusting resources based on demand, thus eliminating the costs associated with unused resources and reducing the need for manual adjustments during traffic spikes. Users can configure autoscaling through simple controls, either via the command line interface or the control panel, and it can be set to react to various metrics, such as requests per second, CPU and memory usage, concurrent connections, and the 95th percentile response time. The platform checks deployments every 15 seconds and utilizes a 1-minute stabilization period to maintain stability after scaling actions. Autoscaling decisions are made on a per-region basis, ensuring that spikes in one region do not affect others, and billing is based on the number of instances and their running time across all regions. The release is part of a series of updates during Koyeb's launch week, and the company encourages users to engage with its community and explore further documentation for more detailed guidance.