We're making checkout more resilient when payment gateways have a momentary hiccup, so customers are less likely to see a payment failure when their payment could actually have gone through fine. This work covers: Automatically retrying transient gateway errors (brief timeouts or service unavailability from providers like Worldline, Zip, Klarna, and Paymark) rather than surfacing them as a hard failure Consistent error handling across all payment providers, replacing ad-hoc logic that behaved differently per gateway Removing avoidable customer-facing failures wherever the platform can safely recover This reduces the risk of customers seeing a failure message on a payment that succeeded, and cuts down on unnecessary "try again" loops at checkout.