Gajendra Babu Thokala
The Data Platform Imperative for AI
Abstract:
As AI becomes embedded in real-world products and operational systems, their limitations are increasingly shaped by the infrastructure responsible for supplying data. Many AI initiatives in the Industry today struggle to deliver timely and reliable outcomes because the underlying platforms were never built for continuous, high-fidelity data movement.
This keynote explores the engineering foundations required to support intelligence at scale. It looks at how architectural decisions around data ingestion, processing, and coordination directly affect system responsiveness, reliability, and the quality of downstream decisions. By contrasting legacy, periodic processing approaches with modern, continuously operating architectures, the talk shows how platforms designed for constant change enable faster adaptation and more consistent behavior in AI-enabled systems.
Rather than concentrating on model capabilities, the discussion emphasizes the engineering required to build data platforms that remain effective as information evolves. The keynote outlines how event-driven, scalable architectures transform delayed insights into continuously updated intelligence, with practical guidance for teams shaping future AI-enabled platforms.
Profile:
Gajendra Babu Thokala is an engineering leader specializing in large-scale data engineering, real-time streaming systems, data governance and AI-driven platforms. With experience across multiple countries including India, Singapore, and the United States, he has led the design and scaling of high-throughput systems that power reliable, intelligent applications at scale. His work has been recognized through professional fellowships and memberships, including BCS Fellowship status, and through his contributions as an author, reviewer, and keynote speaker at international forums. Gajendra is known for turning complex technical challenges into practical, scalable solutions and for sharing real-world insights on building trustworthy real-time systems.