Joshua Hubbard · TAE40116 · National Training Manager at an Australian RTO

Instructional design and AI, for Australian RTOs.

Learner guides, assessments, compliance mapping and LMS-ready content. Built with AI, reviewed and finished by a practitioner, and signed off by your SME before delivery.

Name a unit you have been meaning to redevelop and I will build part of it, free. It is faster than explaining what I do.

10-stage
Pipeline
1 unit
Per day
14+
Deliverables per unit
100%
Standards mapped
Learner guide page
Learner Guide
Simulated workplace document page
Workplace Docs
Assessment workbook page
Assessment

Four reasons RTOs call me.

You don’t have a content problem. You have one of these – and content is how I fix it. Whichever it is, I start the same way: you name a unit, I build part of it, and you judge the work before anyone talks about scope.

Audit exposure

Gaps between your content and the standards that an ASQA visit will find. You need them closed before the visit, not after.

Transition deadlines

A training package is changing and the clock is on. Every affected unit needs rebuilding against the new source document before your window closes.

No one to build it

A scope extension stuck at “almost ready” for months. A qualification your trainers don’t have capacity for. A unit that keeps getting pushed. The bottleneck is always the same: production isn’t moving because there’s no one inside to move it.

AI projects in your RTO

You’re already using AI tools, or being asked to. Assessment marking support, LMS AI features, building your team’s own content workflows. You need someone who understands how AI actually works and what VET compliance requires of it.

start here • one question, then I build
Which unit have you been meaning to redevelop, and never got to?
The superseded one. The one you inherited. The one where the assessments have never quite sat right. Every RTO has one sitting on the list.
Tell me which it is and I’ll build part of it, properly, and send it to you. Then you can judge the work instead of reading claims about it.
what you send
A unit code
Or just a description if the code isn’t handy. Tell me what’s wrong with the current version and I’ll aim the sample at that.
what you get back
A few pages, built
Learner guide or assessment, whichever is more use to you. Real output from the production system, in your context. Not a brochure.
Already have materials for it? Built internally, by an agency, or with AI. Send those instead and I’ll run them through the production system, then show you what it finds in your own materials. Same deal, no cost.
Send me a unit →

See Learnbuilt in action.

Why Learnbuilt exists, and what makes the output hold up under scrutiny. From a practitioner who runs it every day inside a working Australian RTO.

Ten stages. One unit. In front of you.

A real run of the production pipeline – from training-package analysis to a branded, QA-passed pack. This is the engine behind every build.

learnbuilt — pipeline run

It’s not priced by the hour. It’s priced by what it gives back.

Every unit returns 40–100 hours your team would otherwise spend. I don’t bill that time – I give it back to you. What you pay for is the system, the standards expertise, and the judgment that keeps your materials current against the standards.

The materials are the easy part. What your team does with the time back is the point.

01
Most RTOs don’t have a dedicated team.

Your training staff are stretched across multiple qualifications at once. The pitch isn’t “cut headcount.” It’s what does your team do with the time back?

02
A trainer with 40–100 hours back is a different trainer.

They’re coaching. Adapting to individual learners. Building the relationships that make the difference between someone who finishes and someone who doesn’t.

40–100 hrs returned per unit
03
The production argument is almost beside the point.

A typical internal designer takes 4–5 days per unit. That’s before they start on images, video, or a branded LMS. Learnbuilt produces the full text, assessments, content images, and Canvas HTML for each unit, complete and standards-mapped, ready for your SME’s review. But the real number is what happens after that.

Full stack · standards-mapped · LMS-ready

Purpose-built for Australian VET.

Other tools produce content. Learnbuilt produces the full evidence stack your SME validates.

Other AI tools Internal team External LD agency Learnbuilt
The numbers
Per-unit cost Subscription ~$50–200/month, output incomplete $7,900–$10,400/unit all-in (salary, super, leave, tooling) ~$1,000/unit – learner guide and mapping only a fraction of in-house – full stack included, scoped in your proposal
Turnaround Hours – but what you get is incomplete 4–12 weeks per unit 2–4 weeks per unit Days, not months
Your team’s role Prompt, edit, and finish it themselves They build everything You brief, manage, and chase Review and sign off
What’s included
Full learner guide Partial – text only, no AQF calibration – 4–5 days of work ✓ 8,000–15,000 words, AQF-calibrated
Assessment workbooks – additional days Sometimes – extra cost ✓ Knowledge + 1–2 project assessments
Simulated workplace documents – rarely produced ✓ Branded to your simorg
Compliance mapping matrix Manual – time-intensive ✓ Every PC, PE, KE & FS traced
QA report ✓ Mapped to the unit requirements
Canvas / LMS-ready HTML – manual upload ✓ Upload and go
Quality signals
Standards mapping – none Manual – as good as the person running them Varies ✓ Codified – same rules, every unit
AQF level calibration – generic output Depends on the writer Varies ✓ Built into every stage
Video / audio evidence check – not assessed Ad hoc Varies ✓ Multi-layered check, every unit
Human moderation gate – AI output is the final output Varies ✓ Expert human review, every unit
Australian content & legislation – US-centric, wrong spelling Varies ✓ AU legislation, AU industry examples

Your L&D staff are probably already prompting AI. So am I.

The difference is what governs the output.

Without the orchestration layer (AQF level calibration, simulation classification, compliance mapping, and QA against the unit requirements), prompting produces content that reads well but may not hold up under scrutiny. Learnbuilt’s system runs the same QA checks on every unit, every time. The same decision rules fire whether it’s your first unit or your fiftieth.

That’s not a better prompt. That’s a different system.

The maths works. But the real argument isn’t cost – it’s what your trainers do with the time back.

What’s actually inside the system.

Most AI tools give you one layer. Learnbuilt runs three.

01The Compliance Architecture Foundation

The rules that make the output trustworthy

Every unit begins with training package parsing. Elements, performance criteria, performance evidence, knowledge evidence, and assessment conditions are extracted and structured before any writing begins.

From there, AQF level calibration rules determine the cognitive demand required at each level. Simulation classification logic determines how workplace context must be established. Stage 07 checks coverage, compliance mapping and assessment sufficiency against the unit requirements before formatting proceeds.

This is the layer generic AI tools don’t have. It’s why their output reads well but doesn’t hold up when someone with VET knowledge reads it closely.

02Expert Orchestration The moat

The decisions that make the output yours

Orchestration is where practitioner judgment is applied, and where eleven years of working inside Australian VET is baked into the architecture.

Your simulated organisation is configured once and held consistent across every unit in your qualification. Your learner profile shapes reading level, assumed prior knowledge, and the complexity of assessment tasks. Your assessment model determines structure, submission method, and third-party arrangements.

Two RTOs delivering the same unit get materials that read differently, because the configuration is different. That’s not a feature of the generation. It’s a feature of the layer above it.

Human review at Stage 07 checks every unit against the unit requirements and the Standards for RTOs 2025 before formatting begins. Nothing progresses past that gate on the strength of generation alone.

03Production Engine The output

The part everyone fixates on

Yes. AI generates the content. Learner guides, assessment workbooks, assessor guides, mapping matrices, simorg documents, QA reports, LMS-ready HTML. Fourteen deliverables per unit, built to your brand, ready to upload.

It’s consistent, standards-trained, and built around the way VET practitioners think. The output reflects the compliance depth in every decision the system applies.

But it’s the last layer in the system, not the first. The compliance architecture and the orchestration layer run before a single word is written. That sequence is what makes the output deployable, not just readable.

Less than 40% of this system is AI generation. The rest is compliance architecture and expert calibration. That’s why it holds up under scrutiny.

Send me a unit → See the 10-stage pipeline →

Preview on your LMS.

Every unit is built to work on any platform that accepts HTML, and delivered directly on your LMS.

// LMS previews are optimised for desktop. Open on a larger screen for the best experience.

Every unit includes all of this.

No extras. No add-on fees. The full stack, delivered in one go.

Learner Guide
8,000–15,000 words. 5+ modules. AQF-calibrated language. Australian industry contexts throughout. Your simorg and learner profile applied throughout.
Stage 04 › DOCX + HTML
Assessment Workbooks
Knowledge assessment (Assessment 1) plus 1–2 project assessments (Assessments 2–3). Instruments matched to unit type. Conditions clearly stated. All using your simorg.
Stage 06 › DOCX
Assessor Guide
Marking rubrics, model answers, reasonable adjustment notes, and assessor instructions for every assessment task. Your trainers can confidently mark without referring back to the unit descriptor.
Stage 06 › DOCX
Compliance Mapping Matrix
Every PC, PE, KE, and Foundation Skill traced to where it appears in content and how it’s assessed. The document your compliance team and auditors reach for first.
Stage 06 › DOCX
Simulated Workplace Documents
Every policy, procedure, form, register, email, or report referenced in an assessment task. Branded to your simorg, not generic templates. Created specifically for this unit’s tasks.
Stage 06 › DOCX (simorg pack)
QA Report
PASS/FAIL checklist mapped to ASQA 2025 Outcome Standards. Shows your subject-matter expert exactly what was checked and what evidence exists. The documented basis for their sign-off.
Stage 07 › DOCX
Canvas/LMS-ready HTML
Every module as a formatted HTML page, ready to paste into Canvas or your LMS. Branded headers, structured content, embedded activities. No reformatting required after upload.
Stage 09 › HTML (post sign-off)
Simorg Setup (first engagement)
Your Starter Project builds the simulated organisation from scratch: advisory firm structure, divisions, staff, and brand documents. Reused across every subsequent unit, growing as the library grows.
Included in your foundation build
All AI costs absorbed
No platform subscriptions or per-unit fees billed to you. The full cost of running the pipeline is included in your project price. You pay one fixed rate; I absorb the rest.
Included in all project pricing

How an engagement works.

I’d rather show you the work than describe it. So I start by building something, then scope the fix once you’ve seen what you’d actually be buying. You don’t pick a price off a list.

  1. 1 Send me a unit Name one you’ve been meaning to redevelop. No call needed to start, and nothing to sign.
  2. 2 I build part of it A few pages, learner guide or assessment, from the real production system. Free, and yours either way.
  3. 3 A proposal with options If the sample lands, I talk scope, then put options in writing. You pick – no surprises, all deliverables listed.
  4. 4 Build 50/50 terms. Roughly one unit per day, delivered to your Google Drive for your SME to review and sign off.
  5. 5 Ongoing Keep me on as an ongoing support partner, or come back per project. Your call.
Send me a unit →

In-house benchmark: $7,900–$10,400 per unit (all-in salary, super, leave, tooling, and overhead for an L&D team producing 12 units per year, the Australian industry average). A Learnbuilt project comes in at a fraction of that – the exact figure is scoped to your situation, in writing, once I both know what the job actually is.

Two ways to work with me.

I solve the whole RTO workflow problem – not just content. The 10-stage pipeline is my ready-to-deploy engine; I customise it and run only the stages your situation needs. Custom work is scoped to your situation, never off a rate card.

② keep me on the team
Ongoing
I keep working on your operation month to month – assessment QA, compliance admin, LMS updates, content refreshes – from someone who already knows your setup. Scoped to you; your SME keeps sign-off.
How Ongoing works →

Payment: Projects: 50% on commencement, 50% within 7 days of Stage 1–8 delivery. Ongoing: monthly in advance. All prices exclude GST.

Once I have your blueprint, I can build anything.

That first foundation project configures your system once – your brand, your simorg, your assessment approach, your LMS. After that, every future job plugs straight into it. No re-setup. No quoting from scratch. You tell me the problem; I run only the stages that solve it – or work that sits outside the pipeline entirely.

A full new unit

The complete 10-stage pipeline runs, end to end.

Moving to a new LMS

Just the formatting and Canvas-upload stages. Your existing content, migrated and ready to deliver.

Assessments out of date

I rebuild the assessment suite and leave your learner guide untouched.

Delivering face-to-face

Slides and session plans, generated straight off your blueprint – outside the standard pipeline.

Each one is scoped in your proposal – no re-onboarding, because the blueprint is already yours.

ongoing • after a build
Already built with me? Keep me working on it.
The project ends; the monthly work doesn’t. Assessment quality checks, compliance admin, LMS updates, content bursts when a package moves – scoped to your operation, with your SME keeping sign-off.
How Ongoing works →

Answers to what RTOs actually ask.

How is this different from other course-building tools?

Learnbuilt isn’t an AI tool that happened to be applied to VET. It’s a production system built specifically for Australian RTOs, the way a line built for one job outperforms a general-purpose machine every time. Tools like eSkilled AI Creator generate learning-content pages for a global market: broad, multilingual, generic.

Learnbuilt goes into the detail those tools skip entirely, Performance Criteria, Knowledge Evidence, Performance Evidence, Foundation Skills, Assessment Conditions, a multi-layered check that determines whether video or audio evidence is warranted for each unit (and documents the reasoning), and simulated workplace document packs. None of that matters for global content platforms. For Australian RTOs delivering to ASQA, none of it can be skipped.

It also builds to your RTO’s context: your simulated organisation, your learner profile, your industry references, your assessment layout. A generic global tool can’t do that without becoming something else entirely.

I already have an internal training team, why pay you?

Most RTOs don’t have a dedicated team, and when they do, those people are usually stretched across multiple qualifications at once. The pitch here isn’t “cut headcount.” It’s “what does your team do with the time back?” A trainer who isn’t buried in content production has 40–100 extra hours per unit to spend coaching, adapting to individual learners, and building the kind of relationships that make the difference between someone who finishes and someone who doesn’t.

On pure production time: a typical internal designer takes 4–5 days to build one unit, and that’s before images, video, or a branded LMS enter the picture. Learnbuilt produces the full text, assessments, content images, and Canvas HTML for each unit, complete and standards-mapped, ready for your team’s review.

Internal teamExternal LD agencyLearnbuilt
Cost$7,900–$10,400/unit all-in~$1,000/unit, learner guide and mapping onlyA fraction of in-house – assessor guide, simorg docs, Canvas HTML, QA report all included; scoped in your written proposal
Speed4–12 weeks/unit2–4 weeks/unitUp to 1 production day + your review window
Your team’s roleThey build itYou manage themThey review and sign off

The maths works. But the real argument isn’t cost, it’s what your trainers do with the time back.

What happens if the materials fail an ASQA audit?

Learnbuilt handles materials production. You handle validation and the audit relationship. Every materials package is delivered as draft, your subject-matter expert signs off before learner delivery, and your RTO holds the ASQA audit relationship. This is documented in the SOW and applies to every engagement.

If ASQA does find a non-conformance on Learnbuilt-produced materials, the typical audit outcome is a prioritised list of items to remediate within a defined window. I respond fast, most ASQA-prompted updates are scoped and fixed within 1–2 business days at standard revision rates. I’m already across the unit, so the turnaround is much faster than rebuilding from scratch.

Because materials are never published live until your SME signs off, audit findings on Learnbuilt-built units tend to be narrow and specific, not systemic. I don’t put audit-pass guarantees on the site because nobody can honestly make that claim, but I do everything that increases the probability of a clean review, and I respond fast when something needs to change.

Won’t AI-produced materials just sound the same as every other RTO using the same tool?

Only if the production environment is generic, and mine isn’t. Before I write a single line of content, I build a configuration specific to your RTO: your simulated organisation, your assessment model, your learner profile, your brand voice and terminology. Two RTOs delivering the same unit get materials that read differently because the configuration is different.

The compliance layer, Performance Criteria, Knowledge Evidence, Performance Evidence, Assessment Conditions, is the same for every RTO delivering a given unit. It has to be; it comes from the same Training Package document. That part will always look structurally similar. But ASQA doesn’t regulate how you teach, what industry contexts you use, what your simulated workplace looks like, or how your assessment tasks are framed. That’s where your RTO’s identity lives, and it’s entirely configurable.

The most visible differentiator is the simulated organisation. I build it as a professional advisory firm whose clients reflect the industries your learners actually work in. The simorg stays consistent across your entire qualification, same staff, same structure, same documents, but the client scenario changes with each unit. Learners build context over time. No two clients’ simorgs are the same, and because every assessment document is anchored to it, no two clients’ materials read the same either.

Why don’t you just list a price per unit?

Because a unit built from scratch and a compliance rebuild of someone else’s mess are different jobs with the same machinery. A list price would mean overcharging half of you and underquoting the rest. So I build you a sample first, work out what the job actually is, then put options in writing showing exactly what each level of scope costs. You see a number when it’s a real one.

Do I have to commit to ongoing?

No. Projects are one-off – build what you need and I’m done. Ongoing exists for RTOs who’d rather have the recurring work (QA, compliance admin, LMS updates) handled month to month than juggle it in-house. You can start there or come to it later; there’s no requirement to commit.

How does Ongoing work – is it a set number of units?

No. Most of what an RTO needs month to month isn’t new content – units stay current for years. It’s the recurring work around them: assessment quality and marking checks, compliance evidence, LMS updates, and the occasional content refresh when a training package moves. Ongoing is scoped to whichever of those your operation carries. What you’re really buying is that I already know your setup, so a request that would cost a cold provider a week of getting up to speed costs me an afternoon. New builds and migrations are scoped separately as projects, so they get a proper plan and price. You’re never billed by the unit, and never surprised by an invoice.

Is the work mine?

The content I produce is yours – licensed to your RTO for delivery, source files included. The production system stays mine – that’s what I keep improving so your output keeps getting better. If you want to run the pipeline yourself, that’s what an Ongoing partnership is for.

Do you only build course content, or any AI work?

Course content and compliance is the core of what I do, and it’s where the production system really shows. But the same AI capability solves plenty of other RTO problems – automating admin, building internal tools, or untangling a process you suspect AI could handle. If you know AI could fix something in your organisation and it has nothing to do with course content, bring it to a call and I’ll work out whether I’m the right help. These are scoped as their own projects.

Joshua Hubbard, Founder, Learnbuilt
// Learnbuilt Founder, Joshua Hubbard
Built on real VET experience.
Founder, Learnbuilt National Training Manager, Australian RTO TAE40116 8+ Years in Vocational Education
“One unit returns 40–100 hours to your training team. Those hours go back where they belong, with learners.”

I’ve spent 11+ years inside VET, as a trainer, as an assessor, and now as National Training Manager at an Australian RTO. I know what it looks like when a trainer is fully present with a learner. And I know what gets in the way of it.

The trainers who make the real difference are the ones who are in the room, noticing who’s struggling, shifting their approach mid-session, building the kind of trust that makes a learner come back after a setback. Most of them don’t get to do that enough. They’re writing and prompting policies, mapping Performance Criteria, contextualising Word docs that weren’t built for their learners.

One unit produced by Learnbuilt returns 40–100 hours to your training team. Those hours go back where they belong: with learners.

The materials are the starting point. Your practitioners review, validate, and sign off, the compliance responsibility stays with your RTO, as it should. The production work doesn’t have to.

Or just book a call.

Some people would rather talk it through than wait on a sample. Fifteen minutes, no pitch, no obligation – I work out whether your problem is one I can fix, and what it’d take. Most of what I do is VET content and compliance, but plenty of RTOs come to me for AI help that has nothing to do with course content – automations, internal tools, a problem you think AI could solve. Same call.

Book the call
or send an enquiry instead

Already have a unit (built internally, by an agency, or with AI)? Mention it here. I’ll run it through the system beforehand and walk you through exactly what I find.

Something went wrong. Email joshua@learnbuilt.com.au directly.

Response within one business day. joshua@learnbuilt.com.au if you prefer email.

Enquiry received.

I’ll be in touch within one business day.

Writing for RTOs

Honest, practical pieces on VET materials, compliance, and where AI actually helps.

Read the blog →