3rd World Congress on Smart Computing
(WCSC2026)

Organized by  

Soft Computing Research Society

in Association with 

International Auditors for Digital and Data Management Association, Bangkok Thailand

Venue

Novotel Bangkok on Siam Square, Bangkok, Thailand

January 10-11, 2026 

The after-conference proceeding of the WCSC 2026 will be published in Springer Book Series, ‘Lecture Notes in Networks and Systems’.

Sunil Khemka

Beyond the Black Box: Engineering 'Circuit Breakers' and Compliance Guardrails for Autonomous Agentic Workflows

Abstract:

As enterprises transition from passive Generative AI to autonomous Agentic AI, the risk of non-deterministic actions in high-stakes environments (Banking, Healthcare) becomes a critical barrier to adoption. This presentation introduces a novel architectural framework for wrapping autonomous agents with deterministic safety layers. We will explore practical engineering patterns for 'Constitutional AI,' including input/output validation, real-time PII redaction, and 'circuit breakers' that halt autonomy when confidence scores drop below safety thresholds. Drawing from real-world deployments in the financial sector, we will demonstrate how to balance the agility of autonomous agents with the rigorous compliance standards required by global regulators, effectively solving the 'Trust Deficit' in modern AI architecture.

Profile:

Sunil Khemka is a Senior AI/ML Architect and Fellow of the Soft Computing Research Society (SCRS) with over 18 years of experience leading enterprise modernization for Fortune 500 financial institutions. Currently serving as Engineering Manager at Persistent Systems, he specializes in architecting secure, cloud-native AI ecosystems.

A recognized thought leader in AI governance, Sunil serves as Co-Chair of the AI Development Committee at ISAIL (Indian Society of Artificial Intelligence and Law), where he shapes national policy on Deepfakes and AI safety. He is also the Chairman of the AI Advisory Board at Smile Secure and the author of the book "Scaling Trust: AI Platforms for Real-Time Payments, Risk, and Fraud."

His research on Graph Neural Networks, Legacy Modernization, and Agentic AI has been published in IEEE and Springer venues. An inventor with a granted UK Design Patent, Sunil’s work bridges the gap between theoretical AI safety and production-grade reliability. He is a frequent judge for global innovation awards (Smart India Hackathon, BIG Awards) and a mentor to the next generation of technologists.