primis beta update : yesterday we showed the first part of the flow : query rate -> reserve rate now we’re showing you what comes next. your workload runs as usual. usage is emitted. and primis reconciles it against the reserved rate. and the rate matches. this is our core product architecture. query rate -> reserve rate -> attach workload -> reconcile usage. the goal is not to change how builders deploy. the goal is to make compute pricing predicable before the workload runs, and verifiable after it does. we currently have 8 teams/builders trying our closed beta and helping us improving it. we estimate a total of $100k compute volume is running through primis right now. we help these teams save on average around 23.5% of their monthly compute bill and gain tens of hours of productivity. we’re building the pricing layer for compute. our open beta is releasing soon. primis mode