Agentic AI in Platform Engineering
Published by Pearson
Leverage AI agents to reason through platform engineering challenges
- Reduce developer toil, self-optimize pipelines, and enforce governance with AI agents.
- Use AI tools for remediation, observability, and cost optimization.
- Become an AI-enabled platform supervisor and human-in-the-loop engineer.
Agentic AI represents the next frontier of platform engineering, where intelligent, autonomous agents don’t just automate tasks but also reason, adapt, and act on behalf of developers. This course explores how AI-powered platforms reduce cognitive load, accelerate releases, and improve resilience across the SDLC.
By combining engineering platform foundations with autonomous AI capabilities, attendees will see how organizations can evolve into intelligent, self-optimizing ecosystems. This shift isn’t just about efficiency but also about scaling with confidence, enabling innovation, and ensuring governance at enterprise scale.
What you’ll learn and how you can apply it
- Identify where agentic AI enhances developer experience and platform adoption.
- Demonstrate how AI-driven remediation, observability, and governance reduce incidents and manual overhead.
- Evaluate the trade-offs of adopting autonomous agents in CI/CD, operations, and FinOps.
- Envision next steps for introducing agentic AI in your own platform strategy.
This live event is for you because...
You are a platform engineer, DevOps practitioner, or software architect seeking to reduce developer friction, improve system reliability, and explore how AI can elevate your platform beyond automation into self-optimizing, autonomous systems.
Prerequisites
- Basic programming or scripting experience (editing YAML, JSON, or simple Python/JavaScript snippets)
- Basic troubleshooting skills (reading logs, identifying errors, and applying simple fixes)
- Familiarity with CI/CD concepts and version control basics (Git, branches, pull requests)
- Comfortable using a browser-based IDE or terminal
Course Set-up
- GitHub account (free tier), fork repos, run GitHub Actions workflows, and host sample portals with GitHub Pages
- GitHub setup: https://github.com/achankra/agentic-pe-oreilly/blob/main/SETUP.md
- Claude.ai or Claude Code (free tier) for AI-driven troubleshooting, explanations, and interactive agent tasks (browser-only)
- Browser-based development environment: GitHub.dev (lightweight editor) or GitHub Codespaces (optional for terminal access)
- cURL (via Claude Code’s web terminal or browser-based CLI) for simple API/endpoint interactions
Recommended Preparation
- Attend: Introduction to Large Language Models (LLMs) and Prompt Engineering by Sinan Ozdemir
- Attend: AI Agents at Work by Shaun Wassell
Recommended Follow-up
- Read: Effective Platform Engineering by Chankramath, Cheneweth, Oliver & Alvarez
- Watch: Infrastructure & Ops Superstream: Platform Engineering Best Practices
Schedule
The time frames are only estimates and may vary according to how the class is progressing.
Segment 1: Platform Engineering Pain Points & AI Solutions (45 minutes)
- Review manual processes, platform team constraints, and cognitive load
- Introduce Agentic AI: reasoning and adaptation beyond automation
- Compare agent architectures: event-driven vs polling
- Demo: Agent diagnoses failed GitHub Actions workflow
- Mini exercise: Fork a repo, re-run the failing workflow, and use Claude to explain the error and suggest a fix
Q&A (5 minutes)
Break (5 minutes)
Segment 2: AI-Enhanced Development Lifecycle (45 minutes)
- Explore intelligent CI/CD pipelines: self-optimization and automated rollback decisions
- Understand AI-driven remediation patterns and conversational operations interfaces
- Demo: Agent evaluates release readiness using quality gates (test coverage, performance, security)
- Mini exercise: Edit quality gates in repo, rerun the pipeline, and use Claude to generate a release/rollback rationale
Q&A (5 minutes)
Break (5 minutes)
Segment 3: Operational Intelligence & Multi-Agent Coordination (25 minutes)
- Explore AI-enhanced observability: anomaly detection and predictive analytics
- Understand agent communication patterns, conflict resolution, and failure handling
- Demo: Multi-agent simulation (cost optimizer + incident responder) running across GitHub Actions jobs
- Optional interaction: View logs/output in a provided playground, no setup required
Q&A (5 minutes)
Break (10 min)
Segment 4: Implementation Strategy & Organizational Readiness (45 minutes)
- Assess organizational readiness: team skills and infra requirements
- Review phased rollout strategies: monitoring-first evolving to autonomous actions
- Consider risk mitigation: guardrails, ethics, and human oversight frameworks
- Demo: Agent-enabled developer portal hosted via GitHub Pages with AI-driven onboarding/self-service endpoint
- Mini exercise: Trigger the endpoint with curl/browser and use Claude to turn the suggestion into a task list
Q&A (5 minutes)
Break (5 minutes)
Segment 5: From Concepts to Code: Starter Kits in Action (25 minutes)
- Review key architectural patterns and implementation priorities
- Share resources: agent templates and starter repos
- Demo: Showcase starter kit of open-source GitHub Action templates
- Mini exercise (take-home): Fork one template repo and use Claude to scaffold a README with setup and troubleshooting steps
Q&A (5 minutes)
Course wrap-up and next steps (5 minutes)
Your Instructor
Ajay Chankramath
Ajay Chankramath is a coauthor of Effective Platform Engineering and the founder and CEO of Platformetrics, a specialized consulting firm focused on platform engineering and developer experience. With over three decades of leadership experience as chief technology officer, senior vice president, and other senior technology roles at companies like Brillio, Thoughtworks, Oracle, Broadridge, and Xilinx, Ajay is widely recognized as a visionary technologist and has led large-scale digital transformations with a strong emphasis on platform thinking, team enablement, and engineering excellence. Ajay is also a platform engineering ambassador, team topologies advocate, and passionate educator, frequently speaking and teaching on platform strategy, modern software delivery, and technical leadership. He holds a bachelor's degree in computer science, multiple master's degrees in engineering management and computer science, and an MBA. Learn more about his work and insights at platformetrics.com.