AI Agents course

Build intelligent systems that act autonomously

A clear route from zero to expert. Six stages taking you from what AI actually is through to building secure, production-ready agent systems. No prerequisites, just curiosity.
  • FoundationsStart from absolute basics with no coding experience required.
  • SecurityEvery module includes proper security controls and threat awareness.
  • CPD120 CPD credits with certificates and assessments.
CPD hours
AI Agents
120
hours
Hours are fixed by the course design. Timed assessment time is included once on pass.

This course has six stages that build on each other. Move through them at your own pace, but I recommend following the order. Security is woven throughout, not bolted on at the end.

CPD and certificates
If you want CPD evidence and certificates, sign in before you start learning so the system can record your progress. Assessments and certificates help keep this site free for everyone.

⏱️CPD timing

CPD timing

Time estimate (transparent)

I publish time estimates because CPD needs to be defensible. The goal is honesty, not marketing.

Guided learning

120h

Core levels, structured learning

Practice and consolidation

3h

Summary, drills, revisits

Notional range

80 to 180 hours

Quick: core concepts and one exercise per module. Standard: all exercises and reflections for CPD evidence. Deep: extra projects, portfolio artefacts, and peer review.

How I estimate time

I use a notional learning hours approach and I keep the assumptions visible. Where modules are content heavy, I add practice so the hours are earned, not claimed.

  • Reading: 225 words per minute, multiplied by 1.3 for note taking and checking understanding.
  • Labs and practice: about 15 minutes per guided activity, including at least one retry.
  • Reflection for CPD: about 8 minutes per module for a short defensible note and evidence link.
  • Assessments: about 1.4 minutes per question for reading, thinking, and review.

If you study faster or slower, your hours will differ. What matters is that the method is consistent and the activities are real.

🧪Assessment blueprint

Assessment and practice assessment

AI Agents assessment blueprint

Each stage has a timed assessment and a practice assessment. I expand the question bank continuously and keep the marking transparent.

Foundations

mixed

Understand the core ReAct, tool, and memory patterns.

Core concepts

mixed

Reasoning loops, tool contracts, and memory design that stand up under pressure.

Practical building

mixed

Build and orchestrate agents with safe tool use and reliable workflows.

Security and ethics

scenario

Threat modelling, prompt injection defence, and responsible deployment.

Advanced mastery

mixed

Scale, reliability, and evaluation for production grade agent systems.

Capstone

scenario

End to end delivery, governance, and documented operational readiness.

Design rules
  • Every tier must align to behaviour, safety, and observability trade-offs.
  • Questions will reward answers that surface assumptions, risks, and verification plans.

📚Standards and certifications

Standards and certifications

The map we anchor to

I map each course to reputable standards so your learning is defensible at work. I also show common certifications and how their language differs.

Important: This content aligns with these standards and certifications for learning purposes. This is guidance, not endorsement. We are not affiliated with certification providers unless explicitly stated.

Primary anchor standards

  • OWASP Top 10 for LLM Applications 2025
    OWASP

    The definitive guide to security threats in LLM-based systems, including prompt injection, insecure output handling, and supply chain vulnerabilities.

    Official reference
  • OWASP Top 10 for Agentic Applications 2026
    OWASP

    The emerging standard specifically for autonomous AI agents, addressing agentic-specific risks like unbounded tool use, memory poisoning, and multi-agent coordination vulnerabilities.

    Official reference
  • NIST AI Risk Management Framework (AI RMF 1.0)
    NIST

    A practical way to talk about AI risk, governance, and controls across the agent lifecycle.

    Official reference
  • ISO/IEC 42001
    ISO/IEC

    The emerging standard for AI management systems, providing governance structure for AI deployments.

    Official reference
  • EU AI Act
    European Union

    Regulatory framework for AI systems in the EU, classifying AI applications by risk level and mandating compliance requirements.

    Official reference
  • UK NCSC AI Security Guidance
    NCSC

    UK government guidance on securing AI systems, particularly relevant for UK-based deployments.

    Official reference

Certification routes

This course is not endorsed by certification bodies. It is built to prepare you honestly, including where exams simplify reality.

  • NVIDIA Agentic AI Professional
    NVIDIA
    specialist

    The first dedicated certification for AI agent development, covering LLM orchestration, multi-agent systems, and production deployment.

  • Microsoft AI Agents Professional Certificate
    Microsoft/Coursera
    practitioner

    Comprehensive Azure-focused certification covering agent development, deployment, and enterprise integration patterns.

  • Certified AI Security Professional (CAISP)
    Industry
    specialist

    Security-focused certification covering OWASP LLM Top 10, prompt injection defence, and secure AI system design.

  • Microsoft AI-900 and AI-102
    Microsoft
    foundation

    Azure AI concepts provide useful context for cloud-deployed agents and enterprise integration.

  • CompTIA Security+
    CompTIA
    foundation

    Security fundamentals are essential for building secure agents, covering many of the same threat categories.

  • Johns Hopkins Agentic AI Certificate
    Johns Hopkins University
    practitioner

    University-accredited programme covering agent architectures, reinforcement learning, and research-grade implementation.

Organisations and resources

These are the kinds of organisations professionals reference. If you learn how to use them properly, you become harder to mislead.

  • OWASP

    What it is: A global community that publishes practical application security guidance, including LLM-specific threats.

    Why it matters: The OWASP Top 10 for LLM Applications is the closest thing to an industry standard for agent security.

  • Anthropic

    What it is: Creator of Claude and the Model Context Protocol (MCP) standard.

    Why it matters: MCP is becoming a standard for agent-to-tool communication. Understanding it is essential for modern agent development.

  • LangChain and LangGraph

    What it is: Popular frameworks for building AI agents and multi-agent systems.

    Why it matters: Widely used in industry. Understanding these frameworks is practical job preparation.

🧾Terminology translation

Terminology translation

Agent architecture and security

AI agents introduce new vocabulary and new threat categories. Understanding these terms is essential for building and securing agent systems.

Agent, LLM, Chatbot

Plain English

An LLM is the brain. A chatbot uses an LLM to generate text responses. An agent uses an LLM to perceive, reason, and act using tools.

How standards use it

  • Industry practice

    The key distinction is agency: agents can DO things, not just say things.

Common mistake

Treating chatbots and agents as interchangeable. They have fundamentally different risk profiles.

My take

If it cannot call tools, it is a chatbot. If it can modify state or access external systems, it is an agent.

Quick check

What makes an agent different from a chatbot?

Prompt injection (direct and indirect)

Plain English

Direct injection is when a user tricks the agent with malicious input. Indirect injection is when malicious instructions are hidden in content the agent processes.

How standards use it

  • OWASP Top 10 for LLM Applications

    Prompt injection is the number one threat to LLM applications. It cannot be fully prevented, only mitigated.

Common mistake

Thinking prompt injection can be solved with better prompts alone.

My take

Defence in depth is essential. Assume injection will happen and limit the blast radius.

Quick check

Why is indirect prompt injection more dangerous than direct?

ReAct pattern

Plain English

Reasoning and Acting. The agent thinks about what to do, executes a tool, observes the result, and repeats until done.

How standards use it

  • Academic research and industry practice

    The standard pattern for most modern AI agents. Interleaves reasoning with action.

Common mistake

Letting agents skip the observation step, leading to hallucinated results.

My take

If your agent does not ground its answers in tool results, it is just guessing.

Quick check

What are the three steps in the ReAct loop?

🤖Course stages

📦What you will build

Throughout this course, you will create practical artefacts that demonstrate real competence. These are not toy examples but working systems you can show to employers and clients.

🧩Module coverage matrix

Coverage matrix

Module-level coverage

This matrix makes the course defensible: each module is tied to an outcome focus, the anchor standards, and the evidence you can produce.

Artefact templates
LevelModuleOutcome focusDomainsAlignmentAssessmentEvidence
Foundations
1 Understanding Ai
agents-1-1-understanding-ai
Anchors: OWASP Top 10 for LLM Applications 2025, OWASP Top 10 for Agentic Applications 2026, NIST AI Risk Management Framework (AI RMF 1.0), ISO/IEC 42001, EU AI Act, UK NCSC AI Security Guidance
Understanding AI systems and capabilities--Practice + timedTemplate + rubric
Foundations
2 Llms To Agents
agents-1-2-llms-to-agents
Anchors: OWASP Top 10 for LLM Applications 2025, OWASP Top 10 for Agentic Applications 2026, NIST AI Risk Management Framework (AI RMF 1.0), ISO/IEC 42001, EU AI Act, UK NCSC AI Security Guidance
LLM behaviour and limitations--Practice + timedTemplate + rubric
Foundations
3 Command Line
agents-1-3-command-line
Anchors: OWASP Top 10 for LLM Applications 2025, OWASP Top 10 for Agentic Applications 2026, NIST AI Risk Management Framework (AI RMF 1.0), ISO/IEC 42001, EU AI Act, UK NCSC AI Security Guidance
AI agent design, safety, and deployment--Practice + timedTemplate + rubric
Foundations
4 Environment Setup
agents-1-4-environment-setup
Anchors: OWASP Top 10 for LLM Applications 2025, OWASP Top 10 for Agentic Applications 2026, NIST AI Risk Management Framework (AI RMF 1.0), ISO/IEC 42001, EU AI Act, UK NCSC AI Security Guidance
AI agent design, safety, and deployment--Practice + timedTemplate + rubric
Foundations
5 First Ai Interaction
agents-1-5-first-ai-interaction
Anchors: OWASP Top 10 for LLM Applications 2025, OWASP Top 10 for Agentic Applications 2026, NIST AI Risk Management Framework (AI RMF 1.0), ISO/IEC 42001, EU AI Act, UK NCSC AI Security Guidance
AI agent design, safety, and deployment--Practice + timedTemplate + rubric
Core Concepts
1 How Agents Think
agents-2-1-how-agents-think
Anchors: OWASP Top 10 for LLM Applications 2025, OWASP Top 10 for Agentic Applications 2026, NIST AI Risk Management Framework (AI RMF 1.0), ISO/IEC 42001, EU AI Act, UK NCSC AI Security Guidance
AI agent design, safety, and deployment--Practice + timedTemplate + rubric
Core Concepts
2 Tools And Actions
agents-2-2-tools-and-actions
Anchors: OWASP Top 10 for LLM Applications 2025, OWASP Top 10 for Agentic Applications 2026, NIST AI Risk Management Framework (AI RMF 1.0), ISO/IEC 42001, EU AI Act, UK NCSC AI Security Guidance
Tool calling and action execution--Practice + timedTemplate + rubric
Core Concepts
3 Memory And Context
agents-2-3-memory-and-context
Anchors: OWASP Top 10 for LLM Applications 2025, OWASP Top 10 for Agentic Applications 2026, NIST AI Risk Management Framework (AI RMF 1.0), ISO/IEC 42001, EU AI Act, UK NCSC AI Security Guidance
Memory architecture and context management--Practice + timedTemplate + rubric
Core Concepts
4 Design Patterns
agents-2-4-design-patterns
Anchors: OWASP Top 10 for LLM Applications 2025, OWASP Top 10 for Agentic Applications 2026, NIST AI Risk Management Framework (AI RMF 1.0), ISO/IEC 42001, EU AI Act, UK NCSC AI Security Guidance
Agent design patterns and architectures--Practice + timedTemplate + rubric
Core Concepts
5 Architecture Fundamentals
agents-2-5-architecture-fundamentals
Anchors: OWASP Top 10 for LLM Applications 2025, OWASP Top 10 for Agentic Applications 2026, NIST AI Risk Management Framework (AI RMF 1.0), ISO/IEC 42001, EU AI Act, UK NCSC AI Security Guidance
AI agent design, safety, and deployment--Practice + timedTemplate + rubric
Practical Building
1 Building First Agent
agents-3-1-building-first-agent
Anchors: OWASP Top 10 for LLM Applications 2025, OWASP Top 10 for Agentic Applications 2026, NIST AI Risk Management Framework (AI RMF 1.0), ISO/IEC 42001, EU AI Act, UK NCSC AI Security Guidance
AI agent design, safety, and deployment--Practice + timedTemplate + rubric
Practical Building
2 Multi Agent Systems
agents-3-2-multi-agent-systems
Anchors: OWASP Top 10 for LLM Applications 2025, OWASP Top 10 for Agentic Applications 2026, NIST AI Risk Management Framework (AI RMF 1.0), ISO/IEC 42001, EU AI Act, UK NCSC AI Security Guidance
Multi-agent coordination and orchestration--Practice + timedTemplate + rubric
Practical Building
3 Workflow Automation N8n
agents-3-3-workflow-automation-n8n
Anchors: OWASP Top 10 for LLM Applications 2025, OWASP Top 10 for Agentic Applications 2026, NIST AI Risk Management Framework (AI RMF 1.0), ISO/IEC 42001, EU AI Act, UK NCSC AI Security Guidance
Workflow automation and integration--Practice + timedTemplate + rubric
Practical Building
4 Model Context Protocol
agents-3-4-model-context-protocol
Anchors: OWASP Top 10 for LLM Applications 2025, OWASP Top 10 for Agentic Applications 2026, NIST AI Risk Management Framework (AI RMF 1.0), ISO/IEC 42001, EU AI Act, UK NCSC AI Security Guidance
AI agent design, safety, and deployment--Practice + timedTemplate + rubric
Practical Building
5 Integration And Apis
agents-3-5-integration-and-apis
Anchors: OWASP Top 10 for LLM Applications 2025, OWASP Top 10 for Agentic Applications 2026, NIST AI Risk Management Framework (AI RMF 1.0), ISO/IEC 42001, EU AI Act, UK NCSC AI Security Guidance
AI agent design, safety, and deployment--Practice + timedTemplate + rubric
Security Ethics
1 Threat Landscape
agents-4-1-threat-landscape
Anchors: OWASP Top 10 for LLM Applications 2025, OWASP Top 10 for Agentic Applications 2026, NIST AI Risk Management Framework (AI RMF 1.0), ISO/IEC 42001, EU AI Act, UK NCSC AI Security Guidance
AI agent design, safety, and deployment--Practice + timedTemplate + rubric
Security Ethics
2 Secure Implementation
agents-4-2-secure-implementation
Anchors: OWASP Top 10 for LLM Applications 2025, OWASP Top 10 for Agentic Applications 2026, NIST AI Risk Management Framework (AI RMF 1.0), ISO/IEC 42001, EU AI Act, UK NCSC AI Security Guidance
AI agent design, safety, and deployment--Practice + timedTemplate + rubric
Security Ethics
3 Ethics Responsible Ai
agents-4-3-ethics-responsible-ai
Anchors: OWASP Top 10 for LLM Applications 2025, OWASP Top 10 for Agentic Applications 2026, NIST AI Risk Management Framework (AI RMF 1.0), ISO/IEC 42001, EU AI Act, UK NCSC AI Security Guidance
Responsible AI and ethical deployment--Practice + timedTemplate + rubric
Advanced
1 Fine Tuning Models
agents-5-1-fine-tuning-models
Anchors: OWASP Top 10 for LLM Applications 2025, OWASP Top 10 for Agentic Applications 2026, NIST AI Risk Management Framework (AI RMF 1.0), ISO/IEC 42001, EU AI Act, UK NCSC AI Security Guidance
Model customisation and fine-tuning--Practice + timedTemplate + rubric
Advanced
2 Enterprise Architectures
agents-5-2-enterprise-architectures
Anchors: OWASP Top 10 for LLM Applications 2025, OWASP Top 10 for Agentic Applications 2026, NIST AI Risk Management Framework (AI RMF 1.0), ISO/IEC 42001, EU AI Act, UK NCSC AI Security Guidance
Enterprise architecture for agents--Practice + timedTemplate + rubric
Advanced
3 Production Deployment
agents-5-3-production-deployment
Anchors: OWASP Top 10 for LLM Applications 2025, OWASP Top 10 for Agentic Applications 2026, NIST AI Risk Management Framework (AI RMF 1.0), ISO/IEC 42001, EU AI Act, UK NCSC AI Security Guidance
Production deployment and monitoring--Practice + timedTemplate + rubric
Advanced
4 Research Frontiers
agents-5-4-research-frontiers
Anchors: OWASP Top 10 for LLM Applications 2025, OWASP Top 10 for Agentic Applications 2026, NIST AI Risk Management Framework (AI RMF 1.0), ISO/IEC 42001, EU AI Act, UK NCSC AI Security Guidance
AI agent design, safety, and deployment--Practice + timedTemplate + rubric
Capstone
1 Capstone Project
agents-6-1-capstone-project
Anchors: OWASP Top 10 for LLM Applications 2025, OWASP Top 10 for Agentic Applications 2026, NIST AI Risk Management Framework (AI RMF 1.0), ISO/IEC 42001, EU AI Act, UK NCSC AI Security Guidance
Practical application and project work--Practice + timedTemplate + rubric
Capstone
2 Peer Review Certification
agents-6-2-peer-review-certification
Anchors: OWASP Top 10 for LLM Applications 2025, OWASP Top 10 for Agentic Applications 2026, NIST AI Risk Management Framework (AI RMF 1.0), ISO/IEC 42001, EU AI Act, UK NCSC AI Security Guidance
AI agent design, safety, and deployment--Practice + timedTemplate + rubric

🛠️Further practice

Studios and labs to make agent development concrete.

📚How to use this course

  1. Start with Stage 1: Foundations even if you have some experience. The terminology and mental models matter.
  2. Follow the stages in order. Each builds on the previous one.
  3. Complete the practical exercises. Reading alone will not make you competent.
  4. Pay special attention to Stage 4: Security. This is where most real-world projects fail.
  5. Record your learning for CPD evidence. Brief notes after each module work well.
  6. Revisit earlier stages after completing the capstone. You will notice things you missed.

🏁Certification assessment

Each stage has a timed assessment with detailed feedback after submission. You need an account to start. Certificates help your career and help keep this site free.

Practice before the timed attempt
Use the AI Builder Studio, the practice quizzes, and the capstone project to practise first. Then start the timed attempt when you are ready.

📚CPD

Log minutes as you study and practise. Your records stay in this browser. Use the export view when you need a clean summary for your CPD system.

All content is protected. By enrolling, you agree to our terms.

View Course Terms & IP Policy

Quick feedback

Optional. This helps improve accuracy and usefulness. No accounts required.

Rating (optional)

Related Architecture Templates4

Production-ready templates aligned with industry frameworks. Download in multiple formats.

View all

Password and Passphrase Coach

/ai-agents

Foundation

Scores passwords and suggests stronger passphrases.

MFA Method Picker

/ai-agents

Foundation

Chooses MFA methods based on threat fit and device context.

Session and Token Hygiene Checker

/ai-agents

Practitioner

Evaluates session lifetimes, refresh, rotation, and cookie settings.

URL Risk Triage Tool

/ai-agents

Foundation

Checks URLs for risky patterns and produces a quick decision.

Related categories:SecurityIntegrationEmerging