Reliable AI Agents

Chiranjeevisantosh MadugundiChiranjeevisantosh Madugundi is a Principal Software Engineer at Palo Alto Networks, where he architects large-scale enterprise platforms with a focus on reliability, security, and governance. His work spans cloud-native systems, distributed platforms, and AI infrastructure, with emphasis on designing systems that behave predictably and safely in production. He has led the development of mission-critical orchestration and automation platforms, policy-enforced execution layers, and observability solutions used across commercial and federal environments. More recently, he has focused on secure AI platforms and agent-based frameworks that integrate with enterprise data and tooling through controlled execution environments, secure gateways, and auditable policy-driven architectures. Anurag Reddy EkkatiAnurag Reddy Ekkati is a Principal Software Engineer at Palo Alto Networks, specializing in cloud infrastructure security, digital certificate management, observability, and AI-driven cybersecurity automation. He has led the design of secure, scalable platforms that improve system reliability, operational visibility, and enterprise resilience. His work includes building a highly reliable digital certificate management platform that replaced fragmented commercial solutions, eliminated certificate-related outages, and delivered major cost savings. He also designed an internal observability solution that improved metrics, tracing, and logging coverage across complex microservices architectures. Pushpanjali PandeyPushpanjali Pandey is an academician and researcher in Information Technology and Cybersecurity. She is pursuing a Ph.D. in IT Management from Lingayas Vidyapeeth, Faridabad, India, and serves as Director at Gyancity Research Consultancy Pvt. Ltd., where she oversees international conferences and research collaborations. She holds an MBA in Information Technology and a bachelor’s degree in Computer Applications. Her research interests include cybersecurity, digital forensics, cryptography, FPGA-based low-power design, and machine learning. She has published in Scopus-indexed journals and IEEE conferences, has served as Convener and TPC member for international conferences, and brings a strong blend of academic research, industry consultancy, and global academic leadership.Biky ChouhanBiky Chouhan is an Assistant Professor at Chandigarh University, Mohali, and a practitioner-researcher with more than six years of experience spanning artificial intelligence, IoT, Python, blockchain, Linux, and data structures. Mentored by Dr. Bishwajeet Kumar Pandey, he combines rigorous engineering with hands-on pedagogy, with a strong focus on scalable edge-IoT systems, practical blockchain applications, and project-based learning. He holds an MCA from Sikkim University and a BCA from Sikkim Manipal Institute of Technology. Over the course of his career, he has served in academic and training roles at Chandigarh University, GL Bajaj Institute of Technology and Management, NIELIT Delhi and Gangtok, Lovely Professional University, Avhiskaar, Govt. Pharmacy College Sikkim, and in leadership positions for online education initiatives. His applied research centers on secure, efficient, and human-centric systems in healthcare and agritech. He has published 17 patents and has 9 granted patents in areas including smart healthcare wearables, decentralized smart-home control, vibration energy harvesting, intelligent waste segregation, and resilient IoT architectures.

• .- Why Reliability Defines The Next Generation Of AI..- How Agents Are Defined: Code and No-Code Paradigms..- Anatomy of an AI Agent..- Designing Policy-Compliant Agent Flows: Ensuring Goal Alignment and Governance in Execution..- Managing NLP-Induced Instability: When Language Models Challenge Structured Agent Systems..- Prompt Hierarchies, Conflicts and Policy Enforcement..- Orchestrating Multi-Agent Systems with MCP: Reliable Flow Definition and Execution..- Guardrails, Security, and Data Access Control..- From Adversarial ML to Agentic Offense in AI-Enabled Threat Evolution..- Controlled Deployment and Observability in AI Agents: Engineering Transparent Operations..- Building, Evaluating, and Evolving Reliable Agents.