Claude Code meets platform engineering: Securing AI-assisted development workflows

In this session, Manoj and Chandra explore the intersection of AI coding assistants and platform engineering. Tools like Claude Code enable developers to interact with infrastructure through natural language, but this introduces new questions: How do platform teams maintain control? How do they prevent AI agents from bypassing security policies? How do they create productive guardrails?

Key takeaways:
• How AI coding agents like Claude Code change the platform-developer interaction model
• Configuring MCP servers to expose platform capabilities safely to AI assistants
• Implementing permission boundaries and approval workflows for AI-initiated infrastructure changes
• Building golden paths that guide AI assistants toward secure, compliant actions
• Real-world patterns for balancing developer autonomy with platform governance in AI-assisted workflows