Vice President, Cloud Platform and Infrastructure @ Calix | Building Secure Cloud and AI Solutions
Running AI workloads at scale means your infrastructure fails in ways you've never seen before. Token costs spike without warning. Latency compounds across microservices. Your reliability math stops working when AI services hallucinate instead of crashing. Michael Weening calls it "New game. New rules." When you rebuild the platform to support AI-native operations, every assumption about capacity planning, fault tolerance, and cost modeling gets rewritten. Calix's 3rd-generation platform handles this. We designed the cloud infrastructure to absorb unpredictable AI workload patterns while maintaining the reliability our customers expect from the platform. That means building buffering strategies for token consumption, circuit breakers for model inference failures, and cost controls that adapt to variable compute demand. #ConneXions will show what this looks like in production. I'm presenting in three sessions on how the platform actually works. If you're deploying AI features on legacy infrastructure, you're going to hit these problems. Come see how we solved them. https://lnkd.in/gBdJxNUd
Knowing how to win the game. This is why you come. ConneXions 2025 is almost here. Let the game begin. 🔗ow.ly/3cs730sQq8w #ConneXions25
