AI agents are no longer experimental — they’re moving into real-world production systems. But with the ability to reason, call tools, access APIs, and act autonomously comes an entirely new layer of security challenges.

In this session, we’ll explore how to securely deploy AI agents on Google Kubernetes Engine (GKE) while balancing flexibility, scalability, and safety. Learn how to leverage Workload Identity, Network Policies, Secret Manager, and security guardrails to ensure your AI agents remain powerful, reliable, and protected from misuse or vulnerabilities.

Whether you're building intelligent assistants, automation workflows, or autonomous AI systems, this session will help you understand the practical security considerations needed to confidently move AI agents into production.

Agenda

1. Introduction to AI Agents in Production

• Why AI agents are becoming production-ready

• Common architecture patterns and use cases

2. Security Risks of Autonomous AI Agents

• Tool misuse & excessive permissions

• Data access risks and prompt injection concerns

• Attack surfaces in production environments

3. Deploying AI Agents on GKE

• Containerizing AI agents for Kubernetes

• Scaling and orchestration using GKE

• Production deployment architecture

4. Securing AI Agents on GKE

• Using Workload Identity for secure access

• Managing secrets with Secret Manager

• Restricting traffic using Network Policies

• Implementing guardrails and least-privilege access

5. Q&A + Best Practices

• Lessons learned

• Practical recommendations for secure deployment

Event Details

📍 Location: Online (Google Meet)
⏱️ Duration: 45 Minutes
📅 Mode: Virtual Session via Google Meet (Meeting will be shared 1 hour before the scheduled time on your registered email ID)