Code Nexus
CurriculumHow we teachPricingBlogAboutContact
Start learning

How Code Nexus teaches

Watching a solution can feel like knowing it. We ask you to produce one.

Each episode places you inside a fictional team with a practical decision to make. Read the brief, learn the idea, try the task and inspect the result. Later work brings earlier concepts back into use. The research below informs these learning techniques; it does not measure outcomes for Code Nexus as a whole.

Start Episode 1 freeSee the curriculum
Illustrated office scene: a team lead welcomes a new starter beside a desk with the 8-Bit assistant on screen.
Episode 1, Your Access Badge Works!: your first morning on the team.
  • Why it works
  • An episode
  • Readability
  • Story
  • Memory
  • Certifications

The problem with watching

You nod along. Then you meet a blank editor.

Most online tech courses are built around watching. It feels productive, but it builds recognition rather than recall. Decades of research point the same way. People learn more when they do the work, retrieve what they know, and mix topics together.

Students who fail the course

Failure rate. Lower is better.

Traditional lecture33.8%
Active learning21.8%

Freeman et al., 2014. A meta-analysis of 225 studies.

Facts recalled one week later

Share of material recalled. Higher is better.

Re-reading the text40%
Practising recall61%

Roediger & Karpicke, 2006, experiment 2.

Score on a test one week later

Test score. Higher is better.

One topic at a time20%
Topics mixed together63%

Rohrer & Taylor, 2007.

Results from published studies, all drawn on the same 0 to 100 scale. Green is the active approach.

Six principles

Six learning-science findings, and what we did about each.

Learning by doing

Active learning cuts course failure rates by about a third (Freeman et al., 2014).

In Code Nexus: Every scene ends in real work, in one of 28 practical workspaces: a code studio, terminal, source control, cloud console, Docker, Kubernetes, a prompt lab and more.

Situated learning

Skills stick when they are learned in the setting where they are used (Lave and Wenger, 1991).

In Code Nexus: You are the new hire. Tasks arrive as tickets, emails and chat from teammates, the way they do at work.

Retrieval practice

Recalling something beats re-reading it, a week later (Roediger and Karpicke, 2006).

In Code Nexus: Each episode opens with a warm-up drawn from earlier episodes and ends with a scored check.

Spaced repetition

Spreading practice out beats cramming, across 317 experiments (Cepeda et al., 2006).

In Code Nexus: The daily review brings questions back 1 to 2, 3 to 5, 7 to 10, 14 to 20 and 30 to 45 days after you finish an episode.

Worked examples and scaffolding

New learners do better with guided examples that fade as they gain skill (Sweller, 1988).

In Code Nexus: A key idea, a worked example and trap warnings come before each task. Hints unlock one at a time.

Feedback that explains

Feedback with an explanation beats a bare right or wrong (Butler, Karpicke and Roediger, 2008).

In Code Nexus: A wrong answer gets a reason specific to the option you chose, not just a red cross.

Which study techniques actually work?

A major review rated ten popular study techniques. Highlighting and re-reading scored low (sorry, highlighters). The ones at the top are the ones Code Nexus is built from.

High utility

Works across learners, topics and ages

  • Practice testing (used in Code Nexus)Warm-ups, episode checks, daily review
  • Distributed practice (used in Code Nexus)Daily review at growing gaps

Moderate utility

Helps, with limits

  • Interleaved practice (used in Code Nexus)Reviews mix earlier episodes
  • Self-explanation (used in Code Nexus)Debrief: show your reasoning
  • Elaborative interrogation

Low utility

Feels productive, does little

  • Summarisation
  • Highlighting
  • Re-reading
  • Keyword mnemonic
  • Imagery for text
Green means we build it into Code Nexus. Ratings from Dunlosky et al., 2013. Highlighting it yourself does little. Having an expert point at what matters does help. See how we do that

Anatomy of an episode

Nine steps, each there for a reason.

An episode is a small, complete loop. Here is what happens from the first question to the review that finds you again days later, and the principle behind each step.

  1. 1

    Warm-up

    Three questions from earlier episodes the new one builds on.

    Retrieval and mixing

  2. 2

    Story brief

    A teammate hands you a real problem, with a reason to care.

    Situated learning

  3. 3

    Learn in context

    A key idea, a worked example and the traps to watch for.

    Worked examples, low cognitive load

  4. 4

    Predict

    In some coding scenes, guess what the code will do before you run it.

    Pre-testing

  5. 5

    Do the work

    Try the task in a Python workspace or a simulator. Inspect the supplied check and its result.

    Active learning

  6. 6

    Get unstuck

    Hints you reveal one at a time, then a nudge from a teammate.

    Scaffolding

  7. 7

    Pause point

    Episodes split into parts, with a natural place to stop.

    Chunking

  8. 8

    Debrief and check

    Takeaways, then a scored check that explains every answer.

    Feedback

  9. 9

    Days later

    The daily review brings it back, just before you would forget.

    Spacing

One episode from start to finish. The first episodes of a season have less to warm up on, and the predict step appears in a few coding scenes rather than every one.

Designed to be read, not skimmed

We highlight, so you don't have to guess.

Technical text is dense. New learners can't yet tell the key idea from the detail, so they highlight everything, which works out the same as highlighting nothing. We do that work for you. Every note leads with the one idea that matters, marks the terms you need, flags the traps, and walks through a worked example line by line. In research this is called signalling, and it helps most when the material is new to you.

Across the course that means 339 key ideas, 382 trap, rule and tip boxes, 239 worked examples (222 annotated line by line) and a glossary of keyword definitions that appear right where the word does.

medium · taskThe difficulty is labelled before you start.

Reading a number from the user

Key idea

input() always gives you a , even when the user types digits.

When someone types 42, Python receives the text "42", not the number. Text and numbers behave differently, so convert first, using , and calculate after.

Common trap

"5" + 3 raises a TypeError. Python will not guess which one you meant.

Rule of thumb

Convert once, at the moment you read the value, and work with the number from then on.

1age_text = input("Age? ")1
2age = int(age_text)2
3print(age + 1)3
Age? 41
42
  1. 1age_text holds the text "41".
  2. 2int() turns that text into the number 41.
  3. 3Now arithmetic works: 41 + 1.

The conversion sits on its own line, so a failure points straight at it.

On the PCEP-30-02 exam· Certified Entry-Level Python Programmer

  • The exam outline includes input, output and type conversion. Expect to predict what int(input()) does with a given entry.
A sample note, shown with the same component learners see inside an episode. Hover or tab to the underlined words.

  1. Key idea first

    One highlighted sentence before any detail, for the reader who skims.

    Signalling

  2. Keyword identification

    Terms you need get a dotted underline. Hover, focus or tap for a plain-English definition.

    Words next to meaning

  3. Code chips and bold

    Exact names look like code. Emphasis is reserved for the words that carry the point.

    Signalling

  4. Trap, rule and tip boxes

    The common mistake is named before you make it. Never more than two per note.

    Signalling, coherence

  5. Worked example, line by line

    Numbered notes sit beside the code they explain, so you never hunt for the link.

    Worked examples

  6. On the exam

    Where the exam words it differently, the note says how.

    Transfer to the certification

Colour means one thing

Green is right, red is wrong, amber is a trap. Colour always comes with an icon and words.

Underline means a definition or a link

Never decoration, so an underline is always worth reading.

Short parts, natural stops

Episodes split into parts with a pause point between them, so you can stop cleanly.

Built for everyone

Keyboard shortcuts, reduced-motion support, and layouts that hold up on a phone.

Stories and role play

You aren't told about the incident. You're in it.

A story gives every concept a reason to exist. Research on narrative comprehension suggests people follow and remember connected stories more easily than a list of facts. In Code Nexus, the story is the work.

You play a junior engineer. Teammates have distinct personalities, ask for decisions and push back. That is role play: it moves you from knowing about a skill to using it, the way simulation training does in medicine and aviation.

Motivation research names competence, autonomy and relatedness as its drivers (Ryan and Deci, 2000). The cast is the relatedness: people you want to help. The seasons are a career arc, from your first day in IT to releasing AI systems, and each season finale is a boss scene with an optional timer that never fails you.

Three engineers in a control room with two red emergency buttons on the desk and the 8-Bit assistant on the wall screen.
Season 8 finale, Creator Night: The Bill and the Breach. Two alarms, one team, your call.

Team chat, a sample

Kabelo Mokoena

so the bill doubled and a storage bucket is public. both. at the same time.

Priya Naidoo

Cost is my first question. Who can read it is Lydia's. Decide the order before you touch anything.

Maya Chen

Contain first, then count. You make the call. I will ask you why.

Meet the team you'll work with

  • Maya Chen
  • Kabelo Mokoena
  • Amara Okafor
  • Priya Naidoo
  • Lydia Roe
  • Leo Martins
  • Jax Rivera
  • 8-BIT
Read about the cast
An engineer looks puzzled at a laptop showing a red error while a colleague reviews dashboards.
Season 7, It Works on My Laptop: the build is broken, and that is where the lesson starts.

Failure is part of the lesson

Getting it wrong is data, not a verdict.

Research on productive failure finds that struggling with a problem first, then being taught, leads to deeper understanding than being told the answer up front (Sinha and Kapur, 2021). So we let you try, and we make sure being stuck never means being alone.

  1. 1. Specific feedback. A wrong answer explains why that option fails.
  2. 2. A hint ladder. It's dangerous to debug alone: up to three hints, revealed one at a time when you choose. The last never gives away the full solution.
  3. 3. A teammate's nudge. After a second miss, someone from the team steps in with a gentler angle. Nobody sighs.

Remembering, not just finishing

The point isn't to finish an episode. It's to still know it in a month.

Memory fades quickly, then slowly. Each time you recall something just before it slips, it lasts longer. So new episodes open with a short warm-up on what they build on, and a daily review pulls questions from episodes you finished days or weeks ago. A confidence-by-area view then points you to the episode to revisit for your weakest topic. Think Groundhog Day, except each loop gets shorter.

Days a memory lasts before half of it fades

0255075100
1.73.88.4184189
012345

Successful reviews so far (0 is right after you finish the episode)

Illustrative model, not measured data. It builds on the forgetting curve found by Ebbinghaus (1885) and replicated by Murre and Dros (2015), and assumes each review roughly doubles how long a memory lasts. The daily review brings a question back in windows that end on day 2, 5, 10, 20, 45. Real retention varies by person and topic.

Aligned to the certifications employers recognise

A story on the outside. Exam objectives underneath.

Every scene and every question in Code Nexus is mapped to an official objective from the certification it prepares you for. These are exams set by Microsoft, AWS, GitHub, the Linux Foundation, the Python Institute and the Linux Professional Institute, used by employers around the world to check skills. The story is how you learn the skill. The objectives make sure nothing on the exam is left out.

  1. 01

    Start from the official outline

    Each exam's published objectives are stored with the date we checked them and a link to the provider's page.

  2. 02

    Tag every scene and question

    Each one points at an official objective. Our checks reject any tag that is not a real objective.

  3. 03

    Translate the story into exam language

    "On the exam" notes connect what happened in the scene to the words the exam uses.

  4. 04

    Check coverage automatically

    A scanner reports anything taught but not practised, and readiness checkpoints are weighted like the real exam.

From novice to associate, one certification at a time

Your skills build episode by episode. Each badge marks the readiness check for a real exam, in the order you meet them.

NoviceBeginnerPractitionerAssociate
PCEP-30-02E10
PCAP-31-03E16
010-160E19
GH-900E24
AZ-900E28
CLF-C02E32
AI-901E37
AIF-C01E40
AI-103E48
KCNAE56
AZ-400E63

Episode number, from E01 to E70. E15B and E34B bring the authored total to 72. Each badge sits at its readiness check.

The line counts certifications prepared by each episode. The four levels are our own scale for the journey, not an industry standard, and progress varies by learner. Each certification is issued by its provider, not by Code Nexus.
Certification checkpoints in order
CertificationAfter episodeCertifications prepared so far
PCEP-30-02 Certified Entry-Level Python Programmer101
PCAP-31-03 Certified Associate in Python Programming162
010-160 Linux Essentials193
GH-900 GitHub Foundations244
AZ-900 Microsoft Azure Fundamentals285
CLF-C02 AWS Certified Cloud Practitioner326
AI-901 Microsoft Azure AI Fundamentals377
AIF-C01 AWS Certified AI Practitioner408
AI-103 Developing AI Apps and Agents on Azure489
KCNA Kubernetes and Cloud Native Associate5610
AZ-400 Designing and Implementing Microsoft DevOps Solutions6311

Every official objective, mapped

157 of 157 objectives across 11 certifications.

  • PCEP-30-02 · Python Institute

    Certified Entry-Level Python Programmer

    Objectives mapped15 of 15

    77 practice points

  • PCAP-31-03 · Python Institute

    Certified Associate in Python Programming

    Objectives mapped21 of 21

    106 practice points

  • 010-160 · Linux Professional Institute (LPI)

    Linux Essentials

    Objectives mapped19 of 19

    74 practice points

  • GH-900 · GitHub (exam delivered through Microsoft)

    GitHub Foundations

    Objectives mapped8 of 8

    89 practice points

  • AZ-900 · Microsoft

    Microsoft Azure Fundamentals

    Objectives mapped11 of 11

    66 practice points

  • CLF-C02 · Amazon Web Services

    AWS Certified Cloud Practitioner

    Objectives mapped19 of 19

    68 practice points

  • AI-901 · Microsoft

    Microsoft Azure AI Fundamentals

    Objectives mapped7 of 7

    81 practice points

  • AIF-C01 · Amazon Web Services

    AWS Certified AI Practitioner

    Objectives mapped14 of 14

    57 practice points

  • AI-103 · Microsoft

    Developing AI Apps and Agents on Azure

    Objectives mapped14 of 14

    147 practice points

  • KCNA · The Linux Foundation and CNCF

    Kubernetes and Cloud Native Associate

    Objectives mapped13 of 13

    117 practice points

  • AZ-400 · Microsoft

    Designing and Implementing Microsoft DevOps Solutions

    Objectives mapped16 of 16

    147 practice points

Objectives as published by each provider, on the date we last checked them. Mapped means at least one scene or question teaches and practises it. Practice points count the scenes and questions tagged to that exam. Full practice exams exist today for AI-103, AI-901, AIF-C01, AZ-400, AZ-900, CLF-C02, GH-900, KCNA, 010-160, PCAP-31-03, PCEP-30-02.

The material is written by educators who hold Cisco, Microsoft and CompTIA instructor certifications. Meet the team.

Code Nexus is an independent training provider. It is not affiliated with or endorsed by the certification providers named on this page, and prepares learners for exams that each provider issues. Exam names and badges are trademarks of their owners: Python Institute (Open Education and Development Group); Linux Professional Institute; GitHub, Inc. (Microsoft); Microsoft Corporation; Amazon Web Services, Inc; The Linux Foundation.

Built with care, by the numbers

Designed by people who teach for a living.

The team includes Cisco Certified Systems Instructors, Microsoft Certified Trainers and CompTIA-certified educators, and the course was shaped with young learners, working professionals, part-time learners and career returners.

Authored episodes
72
Hands-on scenes
379
Practical workspaces
28
Check questions
705
Hints written
978
Exam objectives mapped
157

Counted from the authored curriculum when this page is built.

Research we built on
  • Freeman, S. et al. (2014). Active learning increases student performance in science, engineering, and mathematics. PNAS 111(23).
  • Roediger, H. L. & Karpicke, J. D. (2006). Test-enhanced learning: taking memory tests improves long-term retention. Psychological Science 17(3).
  • Rohrer, D. & Taylor, K. (2007). The shuffling of mathematics problems improves learning. Instructional Science 35.
  • Cepeda, N. J. et al. (2006). Distributed practice in verbal recall tasks: a review and quantitative synthesis. Psychological Bulletin 132(3).
  • Dunlosky, J. et al. (2013). Improving students' learning with effective learning techniques. Psychological Science in the Public Interest 14(1).
  • Butler, A. C., Karpicke, J. D. & Roediger, H. L. (2008). Correcting a metacognitive error: feedback increases retention of low-confidence correct responses. JEP: Learning, Memory, and Cognition 34(4).
  • Richland, L. E., Kornell, N. & Kao, L. S. (2009). The pretesting effect: do unsuccessful retrieval attempts enhance learning? JEP: Applied 15(3).
  • Sinha, T. & Kapur, M. (2021). When problem solving followed by instruction works: evidence for productive failure. Review of Educational Research 91(5).
  • Sweller, J. (1988). Cognitive load during problem solving: effects on learning. Cognitive Science 12(2).
  • Lave, J. & Wenger, E. (1991). Situated Learning: Legitimate Peripheral Participation. Cambridge University Press.
  • Ryan, R. M. & Deci, E. L. (2000). Self-determination theory and the facilitation of intrinsic motivation, social development, and well-being. American Psychologist 55(1).
  • Guo, P. J., Kim, J. & Rubin, R. (2014). How video production affects student engagement: an empirical study of MOOC videos. Learning at Scale 2014.
  • Graesser, A. C., Olde, B. & Klettke, B. (2002). How does the mind construct and represent stories? In Narrative Impact: Social and Cognitive Foundations.
  • Mayer, R. E. (2009). Multimedia Learning (2nd ed.). Cambridge University Press. Signalling, contiguity, coherence and segmenting principles.
  • Schneider, S., Beege, M., Nebel, S. & Rey, G. D. (2018). A meta-analysis of how signaling affects learning with media. Educational Research Review 23.
  • Murre, J. M. J. & Dros, J. (2015). Replication and analysis of Ebbinghaus' forgetting curve. PLOS ONE 10(7).

Episode 1 is ready.

Stop watching. Start doing.

The first 15 episodes are free. No setup, no credit card, and no need to arrive already knowing the answers.

Code Nexus prepares you for external exams. Each certification is issued by its provider.

Create your free account See pricing
Code Nexus

Practice first. Improvise less later.

We post practical tech tips and the odd 8-BIT opinion. Mostly the tips.

The CPD Group Approved Provider #791172

Learn

  • Curriculum
  • Pricing
  • Create account
  • Sign in

Support

  • Blog
  • Certification guides
  • About Code Nexus
  • How we teach
  • Human help

Legal

  • Terms and Conditions
  • Privacy Policy
  • Cookie Policy
  • Refund and Cancellation
  • Contact: contact@codenexus.co.za

© 2026 Code Nexus. All rights reserved.

Payments secured by Payfast · Billed in ZAR · South Africa