Agentic AI for Platform Engineering: Building Agent-Ready Platforms
AI agents are becoming users of engineering platforms, not just assistants used by developers.
In this hands-on Agentic AI for Platform Engineering workshop, you’ll learn how to evolve an existing Internal Developer Platform (IDP) so AI agents can safely interact with infrastructure, services, environments, operational tooling, and platform workflows.
The focus is not on giving agents unrestricted access. Instead, you’ll work through the platform capabilities required for safe production use: identity, platform context, scoped permissions, tool access, guardrails, human oversight, recovery, and observability.
The session combines platform engineering foundations with the architectural and operational patterns required when AI agents become first-class users of the platform alongside engineers.
What you’ll build
Through guided demos and hands-on labs, you’ll explore how to:
Build golden paths that work for both engineers and AI agents
Create a self-documenting platform knowledge layer
Give agents structured access to services, environments, and platform context
Define agent identity and scoped permissions
Secure infrastructure and tool access through controlled gateways
Build an agent-driven operational workflow for alert triage
Add approval gates, human-in-the-loop controls, and recovery mechanisms
Observe agent actions, LLM usage, audit trails, reliability, and cost
Hands-on capstone
You’ll bring the architecture together by working through an AI-agent-driven alert triage workflow, seeing how an agent can investigate an operational issue, use approved platform capabilities, and take action while remaining inside defined security and governance boundaries.
What you’ll walk away with
By the end of the workshop, you’ll have:
A practical model for treating AI agents as platform users
A pattern for evolving an existing IDP into an agent-ready platform
Approaches for agent identity, context, permissions, and controlled tool access
Guardrail patterns for approvals, security boundaries, and recovery
A framework for AI-agent and LLM observability
A clearer understanding of where agentic platform architectures break in production
A roadmap for introducing AI agents incrementally and safely into platform workflows
Agenda
Agents as Platform Users
What changes when AI agents become users of the platform alongside engineers?
Platform Foundations and Golden Paths
Build the base platform experience that both humans and agents can consume.
Self-Documenting Platform Knowledge
Give agents the structured context they need to reason about services, environments, ownership, and workflows.
Agent Identity, Permissions and Tool Access
Control what agents can see, invoke, and change across the platform.
Hands-on Capstone: Agent-Driven Alert Triage
Use an AI agent to investigate an operational issue and interact with approved platform capabilities.
Guardrails, Security and Recovery
Add approval gates, human oversight, security boundaries, and failure recovery.
LLM and Agent Observability
Track actions, reliability, audit trails, token usage, and operational cost.
Where It Breaks + Q&A
Discuss production failure modes, trade-offs, and implementation lessons.
Who should attend
This workshop is designed for:
Platform Engineers, SREs, DevOps Engineers, Cloud Engineers, Infrastructure Engineers, Platform Architects, Staff Engineers, Principal Engineers, and Engineering Leads working with Internal Developer Platforms, Kubernetes, cloud infrastructure, or engineering automation.
Level
Intermediate to Advanced
Attendees should already be familiar with modern infrastructure and delivery concepts such as cloud platforms, Kubernetes, CI/CD, Git-based workflows, or developer self-service.
Deep machine learning knowledge is not required.
Workshop format
Hands-on + live demos + production architecture discussion
The session is designed around practical implementation rather than high-level AI theory, with a strong focus on how platform teams can safely operationalize AI agents.
Core takeaway
Build a secure, observable, agent-ready Internal Developer Platform where AI agents can access context, tools, and infrastructure through controlled, governed workflows.