Sibasis

Title: Agentic AI for Microservices: Autonomous Optimization of High-Volume Financial Transactions
This session explores how Deep Q-Network (DQN)-based reinforcement learning agents can autonomously manage transaction routing, predictive caching, and scaling in cloud-native microservices without human intervention or predefined rules. 

Attendees will gain insights into how individual agents learn signals like server load, latency, and cache hits to dynamically optimize system behavior. We’ll examine benchmark results comparing DQN against traditional Round-Robin, showing significant gains in response time, resource efficiency, and stability. The session also walks through implementation tips using Python and open-source frameworks.

The talk is designed for students, researchers, and professionals interested in AI-powered DevOps, autonomous infrastructure, and high-performance computing. We will conclude with a discussion on future directions such as multi-agent learning, Kubernetes integration, and compliance-aware decision-making in real-time systems.

Biography:

Sibasis Padhi is a Staff Software Engineer at Walmart Inc. and an IEEE Senior Member, an expert in Microservices & Cloud Performance Optimization for FinTech Solutions. With over 18 years of experience, he has led high-impact engineering initiatives that optimize transaction systems at scale using AI, caching, and distributed systems. His work has been presented at the IEEE Cloud Summit 2025 and featured in peer-reviewed publications. Sibasis is also a speaker, mentor, and technical reviewer, with active contributions to open-source and AI-driven microservices architectures.

Email: