Satish Bhambri

From Transformers to Agentic Intelligence: How RAG Systems Are Evolving

Abstract:

The rapid evolution of large language models has transformed how intelligent systems are designed and deployed, moving from static transformer-based architectures toward agentic, retrieval-augmented frameworks capable of reasoning, planning, and acting in real-world environments.
This keynote will trace that journey, from foundational transformer models to modern Agentic RAG systems, highlighting practical design patterns, system architectures, and real-world lessons learned while deploying these systems at scale. Drawing from enterprise AI deployments, regulatory and compliance use cases, and ongoing research, the talk will explore how agentic workflows, tool orchestration, and retrieval strategies are reshaping areas such as compliance automation, recommendation systems, and decision intelligence. The session will conclude with a forward-looking perspective on emerging challenges and opportunities as AI systems move toward greater autonomy, robustness, and alignment with human objectives.

Profile:

Satish Bhambri is a Senior Data Scientist at Walmart Labs, where he plays a critical role in advancing AI and deep learning within the personalization and ranking organization, leading work on RAG-based agents and large-scale recommendation systems. His work spans enterprise AI, conversational systems, and applied machine learning at scale.
Beyond industry, Satish has contributed to computational astrophysics research at Arizona State University’s School of Earth and Space Exploration, collaborating with Dr. Lawrence Krauss, and is the inventor of an AI-powered Smart Grid system for efficient power distribution. He has authored over ten peer-reviewed publications with IEEE and Springer, with research cited by organizations including NASA and Harvard.
He serves as a Distinguished Fellow and Assessor at the Soft Computing Research Society, has judged major international hackathons including Y Combinator, NeurIPS, DeepInventAI, WUST Hackathon, and Innovate 2025, and has been a featured guest on leading AI and MLOps podcasts, where he shares perspectives on enterprise AI systems, emerging research trends, and the future of intelligent systems.