Learning Hub / AI-Assisted Engineering
AI-Assisted Infrastructure Engineering
Practitioner → Advanced14 lessonsAvailable
Practical AI for infrastructure work: agentic coding tools, MCP servers that expose your platform safely, LLM-assisted diagnostics and log analysis, and where AI should not be trusted with production.
You'll meetClaude CodeMCP serversagentsAI PR reviewspec-driven developmentvibe codingk8sgptguardrails
Start lesson 01 →
What you'll be able to do
- Use Claude Code with skills and hooks for infra automation, without vibe coding
- Put AI into PR review and pre-QA test simulation
- Build a read-only MCP server for cluster diagnostics
- Apply LLMs to log analysis and incident triage safely
Before you start
Kubernetes and automation experience.
How it works
Each lesson: plain-language idea → how it really works → hands-on. Each section ends with a cheat sheet & self-check.
Curriculum
Lessons marked “Read” are ready; the rest are on the way.
Modules
- 01The AI-assisted ops landscapeWhat works today, what doesn'tRead →
- 02AI in daily engineering workScripts, runbooks, RCAs and docs, where it saves real hoursRead →
- 03Engineering with AI vs vibe codingSpecs, tests and review first; when fast-and-loose is fineRead →
- 04Claude Code skills & hooksRepeatable, reviewable automationRead →
- 05MCP servers for infrastructureTools, permissions, read-only firstRead →
- 06AI-assisted PR reviewCatch upgrade risk, security issues and drift before mergeRead →
- 07Simulated testing with AIGenerate failure scenarios and upgrade paths to test pre-QARead →
- 08IaC generation & validationGenerate, then verify with policy and testsRead →
- 09k8sgpt & AI diagnosticsExplaining cluster problemsRead →
- 10Log analysis with LLMsSummarising, clustering, anomaly hintsRead →
- 11Anomaly detection on metricsStatistical baselines before MLRead →
- 12Building an ops agentRunbook automation with human approvalRead →
- 13GPU scheduling for AI workloadsDevice plugins, MIG, time-slicingRead →
- 14Guardrails & governanceAudit, blast radius, data privacyRead →
- 📋Cheat sheet & self-checkEvery command from this section on one page, then 42 questions to check yourself.Open →
Real-world scenarios
Work through each one: symptom → misleading signal → evidence → root cause → prevention.
The agent was confident and wrong
Designing verification into AI-assisted workflows.
This site is a public version of my personal engineering knowledge hub. It intentionally excludes confidential company information and internal operational details.