We were thirty minutes into the architecture review. The team had been running the strangler pattern for six weeks. The contracts were stable. The happy path tested clean. Every person in that meeting was ready to sign off.
Then TiQ surfaced a question I hadn't thought to ask: "Has the inter-service latency been validated against the SLA already committed to customers?"
I asked it. The room went quiet. The lead engineer said "...actually, let me check that." We hadn't validated the latency budget across the new service boundary. We already had 99.9% uptime commitments in signed contracts.