AI Agents course

Core path

Your path through this level

Progress saves in this browser and syncs after you sign in

Completed

0 of 5

You will be able to

Build a small workflow that uses at least two tools
Write guardrails for data access and outputs
Test a workflow with realistic cases and record what failed

Pick up where you left off

A minimal agent pipeline Coordination is the hard part Workflows as reliable plumbing A standard way to connect tools Integration is a contract

Optional

Full module map

Use this if you want the shape of the level before you start

Show

Module 3.1A minimal agent pipeline

Start with a small loop you can test. Complexity is earned.

Open

Prerequisites

Core concepts completed or equivalent understanding
Basic confidence with workflow and integration terms

Outcomes

Explain a minimal agent pipeline in your own words and apply it to a realistic scenario.
Start with a small loop you can test. Complexity is earned.
Check the assumption "You can test the loop" and explain what changes if it is false.
Check the assumption "Side effects are limited" and explain what changes if it is false.

Practice

Build a working flow and verify each control point
Document one failure case and one recovery path

Artefact and failure modes

A working build record with setup, results, and unresolved risks
Hidden complexity. Adding tools and memory too early makes failures harder to locate.
No fallback path. If the agent fails, users still need a clear next step.

Module 3.2Coordination is the hard part

Multi agent systems work when each role is narrow and the coordinator can stop conflict.

Open

Prerequisites

Core concepts completed or equivalent understanding
Basic confidence with workflow and integration terms

Outcomes

Explain coordination is the hard part in your own words and apply it to a realistic scenario.
Multi agent systems work when each role is narrow and the coordinator can stop conflict.
Check the assumption "Roles are bounded" and explain what changes if it is false.
Check the assumption "Conflicts are resolved" and explain what changes if it is false.

Practice

Build a working flow and verify each control point
Document one failure case and one recovery path

Artefact and failure modes

A working build record with setup, results, and unresolved risks
Deadlock and loops. Agents can bounce tasks back and forth. Put time and step limits in the supervisor.
Shared state corruption. If workers write inconsistent state, the supervisor makes decisions on nonsense.

Module 3.3Workflows as reliable plumbing

Automation works when steps are observable, idempotent, and safe to retry.

Open

Prerequisites

Core concepts completed or equivalent understanding
Basic confidence with workflow and integration terms

Outcomes

Explain workflows as reliable plumbing in your own words and apply it to a realistic scenario.
Automation works when steps are observable, idempotent, and safe to retry.
Check the assumption "Steps can be retried" and explain what changes if it is false.
Check the assumption "Failures are visible" and explain what changes if it is false.

Practice

Build a working flow and verify each control point
Document one failure case and one recovery path

Artefact and failure modes

A working build record with setup, results, and unresolved risks
Double execution. If retries cause duplicate side effects, you get billing errors and angry users.
Brittle integrations. External APIs change. Design for versioning, timeouts, and backoff.

Module 3.4A standard way to connect tools

MCP is a structured way to expose tools so an agent can use them safely and consistently.

Open

Prerequisites

Core concepts completed or equivalent understanding
Basic confidence with workflow and integration terms

Outcomes

Explain a standard way to connect tools in your own words and apply it to a realistic scenario.
MCP is a structured way to expose tools so an agent can use them safely and consistently.
Check the assumption "Tools are least privilege" and explain what changes if it is false.
Check the assumption "Auth is explicit" and explain what changes if it is false.

Practice

Build a working flow and verify each control point
Document one failure case and one recovery path

Artefact and failure modes

A working build record with setup, results, and unresolved risks
Overexposed capability. If MCP exposes powerful tools without policy, one prompt becomes a breach.
Untrusted tool output. Tool output can contain malicious strings. Validate and sanitise before using it.

Module 3.5Integration is a contract

APIs are promises. Robust agents handle timeouts, errors, and version changes without drama.

Open

Prerequisites

Core concepts completed or equivalent understanding
Basic confidence with workflow and integration terms

Outcomes

Explain integration is a contract in your own words and apply it to a realistic scenario.
APIs are promises. Robust agents handle timeouts, errors, and version changes without drama.
Check the assumption "Failures are expected" and explain what changes if it is false.
Check the assumption "Data is validated" and explain what changes if it is false.

Practice

Build a working flow and verify each control point
Document one failure case and one recovery path

Artefact and failure modes

A working build record with setup, results, and unresolved risks
Rate limit collapse. Without backoff, one busy hour becomes a self inflicted outage.
API drift. A version bump can break behaviour silently. Monitor and test against real responses.

Optional

Planning and evidence

Objectives, timing, and CPD tracking

Show

If you want to start learning now, leave this closed. Come back when you want to plan your practice or keep evidence for CPD. This is guidance and it is not endorsed by awarding bodies. Standards mapping lives on the course overview page.

Learning objectives

What you will be able to do

1. Build a small workflow that uses at least two tools
Workflows show you how agents behave across many steps and tools.
2. Write guardrails for data access and outputs
Guardrails keep the system safe when inputs are messy or hostile.
3. Test a workflow with realistic cases and record what failed
Testing with scenarios reveals failure modes early.
4. Describe how you would monitor success, cost, and drift
You need practical builds to gain confidence and speed.

What comes next

Next we lock it down because security is the fastest way to lose trust.

Continue to Security and ethics

What changes at this level

Level expectations

Each level is independent but clearly deeper than the last. This panel makes the jump explicit.

Assessment intent

Practical building

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

Style

mixed

50 questions

75 min timed

Pass standard

70%

Not externally certified

CPD timing

Practical Building time breakdown

Defensible timing based on page content: reading, labs, checkpoints, and reflection.

Reading

39m

5,962 words × 1.3

Practice

45m

3 × 15m

Checkpoints

35m

7 × 5m

Reflection

40m

5 × 8m

Estimated total

3h 39m

Based on page content

Claimed hours

30h

Includes reattempts + capstone

Claimed hours exceed on-page estimate by ~27h. Gap will be filled with guided practice and assessment-grade work.

CPD tracking

Fixed hours for this level are 30. Timed assessment time is included once on pass.

View in My CPD

Pricing and CPD Sign in to sync progress and unlock assessments

Progress minutes

0.0 hours

Learning objectives

What you will be able to do

1. Build a small workflow that uses at least two tools
Workflows show you how agents behave across many steps and tools.
2. Write guardrails for data access and outputs
Guardrails keep the system safe when inputs are messy or hostile.
3. Test a workflow with realistic cases and record what failed
Testing with scenarios reveals failure modes early.
4. Describe how you would monitor success, cost, and drift
You need practical builds to gain confidence and speed.

What comes next

Next we lock it down because security is the fastest way to lose trust.

Continue to Security and ethics

What changes at this level

Level expectations

Each level is independent but clearly deeper than the last. This panel makes the jump explicit.

Assessment intent

Practical building

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

Style

mixed

50 questions

75 min timed

Pass standard

70%

Not externally certified

Learning contract

Practical Building outcomes

About 30 hours

Read the explanation first, then use the tools to test the idea. Skip any tool that is not useful for your goal.

Build a small workflow that uses at least two tools
Write guardrails for data access and outputs
Test a workflow with realistic cases and record what failed
Describe how you would monitor success, cost, and drift

Loading content...

Next step

Practise this level, then take the timed assessment

I recommend you start with the practice assessment for Practical building. It is not timed and it helps you write a clear CPD reflection before the full assessment.

Questions

Minutes

70%

Pass mark

Practice assessment

Start the practice assessment for Practical building

It is designed for confidence and evidence, and you can retry as often as you need.

Start practice assessment

Full assessment

AI Agents Practical building assessment

This assessment is timed. It is free to take and you can retry as often as you need.

Detailed feedback on every question
Pass evidence recorded in your account on pass
Personalised recommendations on weak areas

You can complete the course while signed out, and your progress saves in this browser. Sign in before assessments so your pass record is attached to your account.

Courses and assessments are free. There is no paywall for the learning path, practice questions, or formal assessments. Optional donations support hosting, maintenance, and ongoing updates.

During timed assessments, copy and the context menu are restricted to reduce casual cheating. Passed assessments are recorded in your account as evidence.

Course materials are protected by intellectual property rights.View terms

Loading lesson...

Core path

Your path through this level

Progress saves in this browser and syncs after you sign in

Completed

0 of 5

You will be able to

Build a small workflow that uses at least two tools
Write guardrails for data access and outputs
Test a workflow with realistic cases and record what failed

Pick up where you left off

A minimal agent pipeline Coordination is the hard part Workflows as reliable plumbing A standard way to connect tools Integration is a contract

Optional

Full module map

Use this if you want the shape of the level before you start

Show

Module 3.1A minimal agent pipeline

Start with a small loop you can test. Complexity is earned.

Open

Prerequisites

Core concepts completed or equivalent understanding
Basic confidence with workflow and integration terms

Outcomes

Explain a minimal agent pipeline in your own words and apply it to a realistic scenario.
Start with a small loop you can test. Complexity is earned.
Check the assumption "You can test the loop" and explain what changes if it is false.
Check the assumption "Side effects are limited" and explain what changes if it is false.

Practice

Build a working flow and verify each control point
Document one failure case and one recovery path

Artefact and failure modes

A working build record with setup, results, and unresolved risks
Hidden complexity. Adding tools and memory too early makes failures harder to locate.
No fallback path. If the agent fails, users still need a clear next step.

Module 3.2Coordination is the hard part

Multi agent systems work when each role is narrow and the coordinator can stop conflict.

Open

Prerequisites

Core concepts completed or equivalent understanding
Basic confidence with workflow and integration terms

Outcomes

Explain coordination is the hard part in your own words and apply it to a realistic scenario.
Multi agent systems work when each role is narrow and the coordinator can stop conflict.
Check the assumption "Roles are bounded" and explain what changes if it is false.
Check the assumption "Conflicts are resolved" and explain what changes if it is false.

Practice

Build a working flow and verify each control point
Document one failure case and one recovery path

Artefact and failure modes

A working build record with setup, results, and unresolved risks
Deadlock and loops. Agents can bounce tasks back and forth. Put time and step limits in the supervisor.
Shared state corruption. If workers write inconsistent state, the supervisor makes decisions on nonsense.

Module 3.3Workflows as reliable plumbing

Automation works when steps are observable, idempotent, and safe to retry.

Open

Prerequisites

Core concepts completed or equivalent understanding
Basic confidence with workflow and integration terms

Outcomes

Explain workflows as reliable plumbing in your own words and apply it to a realistic scenario.
Automation works when steps are observable, idempotent, and safe to retry.
Check the assumption "Steps can be retried" and explain what changes if it is false.
Check the assumption "Failures are visible" and explain what changes if it is false.

Practice

Build a working flow and verify each control point
Document one failure case and one recovery path

Artefact and failure modes

A working build record with setup, results, and unresolved risks
Double execution. If retries cause duplicate side effects, you get billing errors and angry users.
Brittle integrations. External APIs change. Design for versioning, timeouts, and backoff.

Module 3.4A standard way to connect tools

MCP is a structured way to expose tools so an agent can use them safely and consistently.

Open

Prerequisites

Core concepts completed or equivalent understanding
Basic confidence with workflow and integration terms

Outcomes

Explain a standard way to connect tools in your own words and apply it to a realistic scenario.
MCP is a structured way to expose tools so an agent can use them safely and consistently.
Check the assumption "Tools are least privilege" and explain what changes if it is false.
Check the assumption "Auth is explicit" and explain what changes if it is false.

Practice

Build a working flow and verify each control point
Document one failure case and one recovery path

Artefact and failure modes

A working build record with setup, results, and unresolved risks
Overexposed capability. If MCP exposes powerful tools without policy, one prompt becomes a breach.
Untrusted tool output. Tool output can contain malicious strings. Validate and sanitise before using it.

Module 3.5Integration is a contract

APIs are promises. Robust agents handle timeouts, errors, and version changes without drama.

Open

Prerequisites

Core concepts completed or equivalent understanding
Basic confidence with workflow and integration terms

Outcomes

Explain integration is a contract in your own words and apply it to a realistic scenario.
APIs are promises. Robust agents handle timeouts, errors, and version changes without drama.
Check the assumption "Failures are expected" and explain what changes if it is false.
Check the assumption "Data is validated" and explain what changes if it is false.

Practice

Build a working flow and verify each control point
Document one failure case and one recovery path

Artefact and failure modes

A working build record with setup, results, and unresolved risks
Rate limit collapse. Without backoff, one busy hour becomes a self inflicted outage.
API drift. A version bump can break behaviour silently. Monitor and test against real responses.

Optional

Planning and evidence

Objectives, timing, and CPD tracking

Show

Learning objectives

What you will be able to do

1. Build a small workflow that uses at least two tools
Workflows show you how agents behave across many steps and tools.
2. Write guardrails for data access and outputs
Guardrails keep the system safe when inputs are messy or hostile.
3. Test a workflow with realistic cases and record what failed
Testing with scenarios reveals failure modes early.
4. Describe how you would monitor success, cost, and drift
You need practical builds to gain confidence and speed.

What comes next

Next we lock it down because security is the fastest way to lose trust.

Continue to Security and ethics

What changes at this level

Level expectations

Each level is independent but clearly deeper than the last. This panel makes the jump explicit.

Assessment intent

Practical building

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

Style

mixed

50 questions

75 min timed

Pass standard

70%

Not externally certified

CPD timing

Practical Building time breakdown

Defensible timing based on page content: reading, labs, checkpoints, and reflection.

Reading

39m

5,962 words × 1.3

Practice

45m

3 × 15m

Checkpoints

35m

7 × 5m

Reflection

40m

5 × 8m

Estimated total

3h 39m

Based on page content

Claimed hours

30h

Includes reattempts + capstone

Claimed hours exceed on-page estimate by ~27h. Gap will be filled with guided practice and assessment-grade work.

CPD tracking

Fixed hours for this level are 30. Timed assessment time is included once on pass.

View in My CPD

Pricing and CPD Sign in to sync progress and unlock assessments

Progress minutes

0.0 hours

Learning objectives

What you will be able to do

1. Build a small workflow that uses at least two tools
Workflows show you how agents behave across many steps and tools.
2. Write guardrails for data access and outputs
Guardrails keep the system safe when inputs are messy or hostile.
3. Test a workflow with realistic cases and record what failed
Testing with scenarios reveals failure modes early.
4. Describe how you would monitor success, cost, and drift
You need practical builds to gain confidence and speed.

What comes next

Next we lock it down because security is the fastest way to lose trust.

Continue to Security and ethics

What changes at this level

Level expectations

Each level is independent but clearly deeper than the last. This panel makes the jump explicit.

Assessment intent

Practical building

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

Style

mixed

50 questions

75 min timed

Pass standard

70%

Not externally certified

Learning contract

Practical Building outcomes

About 30 hours

Read the explanation first, then use the tools to test the idea. Skip any tool that is not useful for your goal.

Build a small workflow that uses at least two tools
Write guardrails for data access and outputs
Test a workflow with realistic cases and record what failed
Describe how you would monitor success, cost, and drift

Loading content...

Next step

Practise this level, then take the timed assessment

I recommend you start with the practice assessment for Practical building. It is not timed and it helps you write a clear CPD reflection before the full assessment.

Questions

Minutes

70%

Pass mark

Practice assessment

Start the practice assessment for Practical building

It is designed for confidence and evidence, and you can retry as often as you need.

Start practice assessment

Full assessment

AI Agents Practical building assessment

This assessment is timed. It is free to take and you can retry as often as you need.

Detailed feedback on every question
Pass evidence recorded in your account on pass
Personalised recommendations on weak areas

You can complete the course while signed out, and your progress saves in this browser. Sign in before assessments so your pass record is attached to your account.

Courses and assessments are free. There is no paywall for the learning path, practice questions, or formal assessments. Optional donations support hosting, maintenance, and ongoing updates.

During timed assessments, copy and the context menu are restricted to reduce casual cheating. Passed assessments are recorded in your account as evidence.

Course materials are protected by intellectual property rights.View terms

Stage 3. Practical building

What you will be able to do

Level expectations

Practical Building time breakdown

What you will be able to do

Level expectations

Practical Building outcomes

Practise this level, then take the timed assessment

Start the practice assessment for Practical building

AI Agents Practical building assessment

Stage 3. Practical building

What you will be able to do

Level expectations

Practical Building time breakdown

What you will be able to do

Level expectations

Practical Building outcomes

Practise this level, then take the timed assessment

Start the practice assessment for Practical building

AI Agents Practical building assessment