Engineer reports 11x PR increase after switching to AI-native workflow

At Tano, traditional constraints like mandatory code reviews and heavy communication overhead have been removed. Engineers work independently, with one person owning the full stack from design to backend. Code reviews are now handled by automated agents via GitHub Actions, with human reviews becoming optional after an initial onboarding period. Kakkar notes that while this speed brings risks, such as the agent making logical errors in currency calculations, the team accepts the trade-off of fixing issues post-deployment rather than slowing down for manual verification.

The article highlights that this productivity gain is not just about faster typing but a cultural shift. Kakkar argues that pre-AI bottlenecks like peer review and coordination no longer apply when AI handles the bulk of the work. He suggests that for straightforward features, a single engineer using AI is more effective than a small team coordinating manually. However, he warns that this approach requires a high level of trust and can lead to context switching issues if not managed carefully.