Mr. GAJENDRA BABU THOKALA

Title of the Talk:
Beyond the Model: Why Data Infrastructure Decides the Future of AI

Abstract:
The discussion about AI has been mostly centered on models. However, as AI transitions from experimental environments to real world products and operational systems, the ability of AI to perform effectively is being determined by whether or not the data that feeds it continues to flow through the platforms supporting it. While many efforts have failed to deliver on their promises due to poorly performing models, this failure was also caused by the fact that those platforms were originally developed for intermittent, low fidelity (i.e., inaccurate) data movement and not for continuous, high fidelity (i.e., accurate) data movement.The presentation will shift the focus from model performance to an engineering perspective. This presentation will show how design choices related to data ingestion, data processing and data coordination impact system responsiveness and reliability along with the overall quality of decisions made at each level of decision-making in these systems. In comparing periodic process based legacy architecture with the new constantly changing architecture, the speaker will demonstrate that the type of platform that is able to handle constant changes results in AI systems that react much faster to changes than they do today and make consistent decisions relative to past experiences. The presentation will provide actionable advice for engineering leaders who wish to help build the next generation of AI enabled platforms and will clearly convey the message that the future of AI will be determined by the platforms which are used to support them.

Bio:
Gajendra Babu Thokala is a senior technology leader. He has extensive experience in designing scalable high-volume (data) engineering solutions as well as in developing large scale, real-time streaming systems. Gajendra has successfully designed, scaled and delivered numerous high throughput systems to thousands of customers.
He is an IEEE Senior Member and a BCS fellow. He is an author of two books, reviewer and international keynote speaker. He is also well-known for taking complex technical challenges and turning them into working, scalable platforms; and providing first-hand knowledge and insight on how to build trusted systems capable of processing in real-time.