Divyesh Govaerdhanan
Technical Team Lead - Cloud Assert
Herndon, Virginia, United States
Actions
Divyesh Govardhanan is a Microsoft MVP in Azure PaaS and Technical Team Lead at Cloud Assert Technologies with 10 years of experience building secure, scalable enterprise solutions using the .NET ecosystem and Microsoft Azure. A multi-certified Azure professional and CompTIA Security+ holder, Divyesh is passionate about mentoring engineers and contributing to the developer community.
Badges
Area of Expertise
Topics
Zero-Trust for LLM Apps on Azure
LLM apps fail in predictable ways: prompt injection crosses trust boundaries, RAG retrieves the wrong thing (or too much), secrets leak through tooling, and teams discover incidents only after users report them. This session shows how to design an LLM application on Microsoft Azure with a zero-trust mindset, treating prompts, retrieval content, tools, and model outputs as untrusted inputs, and how to translate that into concrete Azure controls.
Zero trust for AI agents: securing Azure AI Foundry workloads in production
AI agents built on Azure AI Foundry introduce a new class of security challenges that traditional perimeter-based thinking does not solve. An agent that can call APIs, query databases, and invoke tools needs a trust model where every action is verified, every credential is scoped, and no component is implicitly trusted, even within your own Azure subscription.
This session walks through a zero trust architecture for production AI agent workloads on Azure AI Foundry. We cover Managed Identity for agent-to-service authentication without secrets, Azure RBAC scoping for tool and data access, network isolation with Private Endpoints and Azure Container Apps ingress policies, prompt injection defense patterns, and audit logging with Azure Monitor.
Real patterns from a production AI-powered application are used throughout, so attendees leave with implementation-ready guidance rather than theory.
Please note that Sessionize is not responsible for the accuracy or validity of the data provided by speakers. If you suspect this profile to be fake or spam, please let us know.
Jump to top