Component · not built
This page is a promise, not a product.
The plan: one "brain" server that watches load across the mesh, predicts what's coming, provisions or retires $6 VPS nodes, and hands each one a role — realtime, edge cache, shard, or storage. None of it exists yet. We're describing it so the waitlist can push on it.
The intended job
- Demand sensing across realtime throughput and shard load
- Predictive provisioning ahead of spikes, not after
- Role assignment so a fresh node knows what to be
- Automatic drain and removal when demand falls
Why it's last
Autoscaling a system that isn't proven under real load is how you automate a mistake. The mesh, transport, streaming, and storage come first and stand on their own. The brain is the convenience layer on top — worth building only once there's demand to point it at.