Work through engineering problems.
Debug a program, investigate a cloud configuration, test an AI assistant or review a software release. Build your skills across coding, infrastructure, AI and delivery.
Designed for complete beginners and career changers, this hands-on AI Engineering course takes you from curious beginner to building practical AI solutions.No experience required. Just curiosity and a healthy suspicion of robots.
The Code Nexus product map
Eight products, one team, and you're on the rota.
No coding experience needed. Start with the foundations and build towards AI. Each season is a group of connected episodes: guided problems you work through in your browser.
Seasons 1–3
Start with computer basics. Use Python, a programming language, to write small tools and keep a history of your work.
Seasons 4–6
Learn about online computing, find information in documents and check AI answers.
Seasons 7–9
Practise checking and sharing software, handling simulated failures, and limiting what an assistant may do.
Start with files, accounts and safe computer habits. Learn Python, a programming language, to calculate values, organise information and handle repeatable tasks.
Use Linux to work with files and commands. Use Git and GitHub to keep earlier versions, explain changes and practise reviewing someone else's work.
Explore cloud services: computing you rent online. Compare running costs, access and responsibilities in Azure and AWS practice scenarios.
Learn when AI is useful, find information in supplied documents and practise the parts of a support assistant: answers, tools, quality checks and safety boundaries.
Learn how software is packaged, checked and released. Work through simulated service failures, updates and recovery decisions.
Explore assistants that take steps towards a goal. Practise tool requests, cooperation between assistants and decisions that need human approval.
Debug a program, investigate a cloud configuration, test an AI assistant or review a software release. Build your skills across coding, infrastructure, AI and delivery.
Choose Core programme for the full programme, or Certification prep for certification practice and study guidance. External exams and vouchers are separate.
Work through episodes at your own pace on a tablet, laptop or desktop with internet access. Examples, hints and feedback help you get started.
Who needs what, by when, and why it matters to the team.
A key idea and a worked example from the teammate who uses it at work.
Use a Python workspace or a simulated terminal, Git workflow or cloud console. Check the supplied result.
Finish with a tested change, a decision record or an incident note you can defend.
Evidence in the real world
Outages, bad inputs and misleading metrics happen at every scale. Code Nexus lets you practise the same kinds of decisions where a mistake costs a retry, not a public incident report.
The Code Nexus team
Seven people and one AI assistant who will question your assumptions, read your evidence and ask what happens in production.

Senior AI Engineer · Your mentor
Calm, precise and quietly dry. Starts you on checklists, then takes them away. "It runs. Now show me what happens when the input is wrong."

Platform and DevOps Engineer
Linux, containers, pipelines and incident recovery. Writes his best runbooks right after something breaks. "It worked on my machine. I've updated the runbook to say why that wasn't enough."

ML and Data Engineer
Data quality, evaluation and responsible AI. Sceptical of any result that looks too good. "The accuracy improved. Which group did we leave out?"

Cloud Architect
Azure, AWS, architecture and resilience. Warm, and suspicious of unexplained spend. "Azure will provision it. Finance will ask whether we needed it."

Security Engineer
Identity, least privilege, secrets and incident control. Deadpan about boundaries. "Nobody will guess the URL is not an access control."

Product Manager
Requirements, priorities and stakeholders. Arrived asking "can we add AI?" and now brings constraints. "I brought a reason and a success measure this time."

Creator Partnerships Lead
The voice of Code Nexus's creators: user impact, accessibility and unusual tickets. "Nobody uses it that way? I have twelve screenshots."
Code Nexus Internal AI Assistant
The team's internal AI assistant. Helpful, fast, and occasionally too sure of himself, which is why his work always comes with something you can verify.
8-BIT has an answer for everything. Checking it is your job.
The Code Nexus curriculum
Start with files, accounts and your first code. Progress towards AI assistants, software delivery and simulated agent workflows. Explore each season for its tasks, tools and practice projects.
Season 1
Get comfortable with files, accounts and error messages, then write your first small programs.
The badge works. Let's see what else does.
Season 2
Use Python to organise information, calculate results and turn repeated steps into reusable code.
Nobody expects the IndentationError.
Season 3
Keep a history of your work, recover earlier versions and practise making changes with a team.
Your code has colleagues now.
Season 4
Understand rented online computing services and compare their costs, access and responsibilities in practice scenarios.
Scales to infinity and beyond. So does the invoice.
Season 5
Understand what AI can and cannot do, check the quality of its results and decide when it is appropriate.
"Can we add AI?" is the start of the meeting, not the end.
Season 6
Practise the parts of a support assistant: finding document passages, checking answers and limiting its actions.
Confidence is not a source.
Season 7
Practise packaging software and keeping a service running through simulated checks and failures.
Kabelo has heard this one before.
Season 8
Work through checking, approving and releasing a change, including what to do when it goes wrong.
The deploy button is not the ending.
Season 9
Explore how AI assistants request tools, share work and ask a human before taking certain actions.
The agents have a plan. Check it.
Practise skills mapped to 11 external certification exams, from PCEP to AZ-400. Season checkpoints show what needs more work. Book each exam with its provider, which issues the certification.
How we map certification objectivesPCEP-30-02
PCAP-31-03
010-160
GH-900
AZ-900
CLF-C02
AI-901
AIF-C01
AI-103
KCNA
AZ-400
Tell us what you have tried. A Code Nexus mentor replies privately and helps you work through the problem.
Tell a mentor which episode has you stuck, then keep the conversation going by message.
Both paid plans include four 30-minute mentor calls each month, subject to availability.

A record of your work
Keep track of what you have completed and what to practise next.
Write a Python tool, fix an access issue or explain a release decision.
Return to the same team as their projects grow and new problems appear.
Find out which skills to revisit before moving on.
Learning by doing
A sourced reference to active learning, retrieval practice and study methods, with a downloadable chart, research context and limits for each finding.
6 min read · Updated
AI engineering
Questions about AI and work, with research sources and practical steps for checking tools and their results as you consider what to learn next.
4 min read · Updated
Learning by doing
Starting with no coding experience? See how to approach an AI engineering course, what equipment you need and how to judge your first practical steps.
4 min read · Updated
You can be curious about AI and worried about it. Understanding what it does, checking its output and knowing when a person must decide are part of the learning.
AI can change tasks and disrupt work. The ILO's 2025 research suggests job transformation is the more likely overall effect, but exposure differs between occupations. It cannot predict what will happen to your job, and this programme cannot guarantee employment or income.
Read the ILO research on AI and workAI can produce misleading answers and be misused. Researchers also study future risks where people could lose control of advanced systems; experts disagree about their likelihood. Checking sources, limiting permissions and requiring human review help address particular risks, but do not solve every safety problem.
Read the 2026 International AI Safety Report summaryPress start.