India’s Scale Demands Planning For Failures Beyond Your Stack: Razorpay’s Prabhu Ram
Executive Take
Most outages blamed on vendors are actually a design failure to plan for them. Leaders should ask engineering teams what happens when a dependency they don't control fails, not just when their own code does.
Executive Summary
At Inc42's CTO Summit 2026, Razorpay SVP of engineering Prabhu Ram said companies must design for failures in infrastructure layers they don't control, not just their own systems. He contrasted this with Apple, Microsoft and Amazon, which own more of their tech stack. Ram also warned AI cannot fix weak engineering or poor observability.
Why It Matters
Technology leaders running high-volume systems like payments or ticketing need resilience plans that cover vendors and cloud layers, not just internal code. CTOs adding AI should know it adds risk on top of weak monitoring, it doesn't fix it.
Bizquad Perspective
The real blind spot isn't technical architecture, it's that most companies never map which of the seven stack layers they actually don't control until something breaks.