What we are funding
Three things: on-premise deployment for teams whose data cannot leave their network, a wider tool registry, and the unglamorous reliability work that does not demo well but decides whether anyone can put an agent on call.
We are not building a model. There are excellent ones and we would like our customers to switch between them freely, which is easier when the execution layer belongs to neither.
What stays the same
Pricing stays usage-based. The free tier stays usable for a real side project rather than a demo. And the changelog stays public, including the incidents.