Learning paths
Agentic AI: from your first agent to production
Read these guides in order to go from what an agent is to running agents in production. Each stage builds on the one before it.
Stage 1: Foundations
- Agentic AI
A practical guide to agentic AI: what AI agents are, how they plan and use tools, the protocols, frameworks and patterns behind them, and safe operation. - ReAct loop
ReAct interleaves reasoning, actions and observations so an AI agent can think, use a tool, read the result and adjust. Learn how the loop works. - Function calling (tool use)
Function calling lets a language model request a tool with structured arguments. Learn how tool schemas, the call loop, parallel calls and validation work. - Structured outputs
Structured outputs make a language model return data that matches a JSON Schema. Learn constrained decoding, schema design, validation and when to use it. - MCP (Model Context Protocol)
MCP is the open standard that connects AI models and agents to tools, data and prompts. Learn how servers, clients, transports and auth work, with examples.
Stage 2: Building agents
- LangGraph
LangGraph builds AI agents and workflows as graphs of nodes, edges and shared state, with checkpoints and human approval. Learn how it works and when to use it. - Claude Agent SDK
The Claude Agent SDK exposes the agent loop, tools, sub-agents, hooks and permissions behind Claude Code for your own agents. Learn when to use it. - Multi-agent systems
Multi-agent systems split work among several AI agents with different roles. Learn the main architectures, when they beat a single agent, and their costs. - Orchestrator-workers
In the orchestrator-workers pattern a lead agent splits a task, sends parts to worker agents in parallel and combines results. Learn when it pays off. - Agent memory
Agent memory lets AI agents keep context within a task and across sessions. Learn short-term and long-term memory, retrieval, summarisation and privacy design.
Stage 3: Running them in production
- Evals
Evals are systematic tests of AI behaviour: datasets, graders and metrics run before and after every change. Learn offline and online evals for agents. - Tracing for AI agents
Tracing records every model call, tool call and handoff in an AI agent run as linked spans. Learn what to capture, OpenTelemetry and how to debug runs. - Guardrails
Guardrails are the checks around an AI model that block unsafe input, invalid output and risky actions. Learn the layers, techniques and how to tune them. - Prompt-injection defense
Prompt injection hides instructions in content an AI agent reads, to hijack its actions. Learn direct and indirect attacks and layered defenses. - Sandboxing for AI agents
Sandboxing runs an AI agent's code, commands and browsing in an isolated environment with limited files, network and credentials. Learn the options. - Human-in-the-loop
Human-in-the-loop designs keep people in control of AI agents through approvals, edits and escalation. Learn where to place checkpoints in agent workflows. - Cost and latency budgets
Cost and latency budgets cap how much time, tokens and money an AI agent may spend per task. Learn how to set and enforce them without losing quality.
How to use this path
Read the steps in order; each guide links onward to related topics, so you can go deeper wherever you need to. If you already know a stage, skip to the next one. All learning paths are listed on Learning paths.