Mayur Madnani

Mayur Madnani

Principal Software Engineer

About

Mayur Madnani is a Principal Software Engineer at Candescent, specializing in large-scale distributed systems, data infrastructure, and applied AI. With over a decade of experience designing and scaling platforms across SAP, Walmart, Intuit, JioHotstar, and Candescent, he focuses on solving complex architectural challenges that empower engineering teams and support mission-critical systems serving millions of users.

His work spans data platforms, AI and Generative AI applications, real-time analytics, ML observability, distributed systems, and developer tooling. He is particularly interested in building self-service platforms and engineering capabilities that enable teams to move faster while maintaining reliability and operational excellence.

Beyond his engineering work, Mayur is a published researcher, patent holder, conference speaker, mentor, and active contributor to the engineering community. He has presented research at IEEE conferences, authored patents in AI and large-scale data systems, participated in industry hackathons, and regularly mentors engineers through technical talks, workshops, and community initiatives. Through his speaking engagements, he shares practical lessons from building production-scale systems and explores how AI, platform engineering, and distributed systems are shaping the future of software engineering.

As Large Language Models (LLMs) evolve from isolated chat interfaces into autonomous agents capable of interacting with external systems, the attack surface has dramatically expanded. We are no longer just trying to prevent an AI from generating toxic text; we are defending highly privileged, transactional pipelines against systemic exploitation.

This session explores the Full Stack of LLM application security. We will begin by demystifying the modern threat landscape, exploring how inherent training biases and mathematical generation methods can be actively weaponized to trigger targeted hallucinations and supply-chain vulnerabilities. From there, we will demonstrate how attackers utilize sophisticated direct and indirect prompt injections to reliably bypass standard semantic guardrails.

Crucially, attendees will see how these front-door exploits pivot directly into the transaction layer, hijacking agent tool access to exfiltrate private data or execute unauthorized actions. Finally, the session will shift from Red Team to Blue Team. We will deconstruct why traditional heuristic guardrails fail and introduce robust structural design patterns that secure AI workflows by design.

Attendees will leave with actionable, framework-agnostic strategies to stress-test their environments and automate the defense of their evolving LLM applications.

Read More →
View All Speakers

We use cookies essential for this site to function well. Please click to help us improve its usefulness with additional cookies. Learn about our use of cookies in our Privacy Policy & Cookies Policy.

Show details