Vinay Soni

Title of the Talk :
Intelligence at Scale, Designing Distributed Systems for AI Workloads

Abstract of Talk:
AI workloads introduce new architectural stress points, data intensity, compute bursts, and unpredictable latency profiles. This keynote walks through how distributed systems must evolve to serve large-scale AI inference and training efficiently. We’ll cover core principles for scalability, fault tolerance, and latency optimization, with real patterns from modern AI infrastructure. Attendees will learn practical design choices for model serving, caching, and observability that keep intelligent systems reliable at scale.