AIOps on AWS / Part 1 · Agents from zero / Decision models for fixed choices Contents Part 1 · Agents from zero Contents of the book Start here The opening pages Introduction: what this course builds free Before you start free Getting a model API key and keeping it safe Part 1 Agents from zero What a language model really does Talking to a model: messages, roles and temperature Answers you can trust in code What an AI agent is free One agent loop, step by step Tools for an AI agent Designing tools and context Write the agent loop yourself The same agent with LangChain's create_agent Give the agent an AWS tool Agent middleware: limits, retries and approval Decision models for fixed choices A cost log for every model call Part 2 LangGraph from zero Why build an agent as a graph LangGraph state and reducers Routing, loops and the recursion limit Hand-offs, parallel work and streaming Durable agents that survive a crash Asking a human with interrupt Timeouts, error handlers and caching Agent memory Deep Agents, a ready-made harness Tracing an agent with LangSmith Part 3 One DevOps agent, done safely and measured The demo shop we break all course coming The Kubernetes agent on the real shop coming The Approval Gate and the Executor coming The Ops Console coming The first attack: an instruction hidden in a log line coming What it means for an incident agent to work coming The evaluation harness coming Fair numbers for an agent coming Part 4 Tools the standard way, MCP What MCP is coming Our tools as an MCP server coming Real MCP servers for Kubernetes, GitHub and Grafana coming Your own MCP server for security scanning coming The MCP gateway coming MCP security coming Part 5 From one agent to a team Multi-agent team patterns coming What a team of agents costs coming Designing the context first coming The Incident Commander agent coming The specialist agents coming The Verifier and the Scribe coming When an agent team goes wrong coming Alert intake and paging a human coming Long-term memory for the team coming Part 6 Production on AWS AWS from zero, cost safety first coming Terraform, practised on the laptop first coming Amazon EKS with one IAM role per agent coming The Cloud agent and the AWS incidents coming Kyverno, Secrets Manager and an audit trail coming OpenTelemetry export inside your own network coming Approving agent changes from Slack coming Part 7 Prove it The operations report coming Red-teaming your own agents coming The final evaluation coming Buy or build coming Capstone: a new fault and a new specialist coming Appendix DevOps basics, if you need them Containers from zero coming Kubernetes pods coming Kubernetes deployments coming Kubernetes services coming Kubernetes health probes coming Kubernetes storage coming Kubernetes autoscaling coming Observability from zero coming CI/CD and GitOps coming Part 1 · Agents from zero
Decision models: a probability for every answer when a step is a fixed choice Using a decision model (TypeSafe's Jev, or OpenAI's Decisions API) for yes or no choices, and why a confidence score is never a safety control.
This page is part of the paid AIOps on AWS course The early-access price is ₹3,333 or $100 , one payment. It rises to ₹7,777 or $177 once every part is released, and buying now gives you every part as it is released, at no extra cost, with lifetime access.
Free to read first: Introduction: what this course builds , Before you start and What an AI agent is .
Already bought it? Sign in to read this page.