Questions, answered straight.

VETos is an AI operating system built specifically for vocational education and training providers. It does two things. It generates training and assessment content (assessment tools, study guides, SCORM packages, training and assessment strategies), and it holds a live regulatory graph that maps your scope, your qualifications and the standards you are held to. The generation side does the work. The graph side makes sure that work lines up with the rules.

RTOs in Australia, and providers, ITOs and PTEs in New Zealand. We also support providers in the UK. It is built for the people who carry the load and the risk: trainers and assessors, learning designers, and the compliance and quality teams who answer to the regulator.

It takes the heavy, repetitive work off your team. Drafting assessment tools against unit requirements. Building study guides and learner content. Producing SCORM packages you can drop straight into your existing systems. Putting together a training and assessment strategy. And it keeps a clear line of sight between what you have produced and the requirements it has to satisfy, so nothing quietly drifts out of alignment.

We're big fans of tools like ChatGPT and Claude — we use them every day, and you should too. This isn't an either/or, and we're not asking you to swim against the tide. The difference is fit for purpose. A general assistant treats every request as a blank page: ask it for an assessment and you get a capable, standalone draft — but one that's disconnected from your scope, from the other resources you've built, and from the standards it has to satisfy, and it won't remember any of it tomorrow.

VETos is the industrial-strength, VET-specific version of that capability. It's built around a centre of record — a live regulatory graph of your scope, qualifications, units and standards, and how they all connect. Everything it generates is anchored to that graph, so your assessment, your study guide and your training and assessment strategy aren't three unrelated files that happen to share a title — they're joined to the same source of truth and to each other. Change a unit on the register and everything grounded in it knows.

It also isn't a single model. VETos orchestrates multiple best-in-breed AI systems — models for reasoning and writing, a graphing layer that holds the relationships, and dedicated image, diagram and (soon) video generation — each doing what it does best and checking the others. That orchestration is what drives accuracy up: every output is cross-checked against the map and the standard, rather than taken on trust from one model's first guess.

So keep using your favourite assistant for everyday drafting and thinking. For the compliance-critical work — content that has to map to a unit and stand up at audit — VETos is the domain-specific layer built for the job: connected, mapped, audit-ready and grounded in your scope, with your people reviewing and signing off.

The foundation is a detailed map of your entire scope. Before VETos generates a single thing, it builds out everything you have the right to deliver: every qualification, every unit or standard, and the requirements sitting underneath them. In New Zealand that is your consent and moderation requirements. In Australia, your scope of registration. Then, and this is the part that does the real work, it captures how all of those pieces connect to each other, not just the list of them.

That connected map is the bedrock of everything that follows. A general AI tool is guessing from whatever it has read on the internet. VETos works from your actual scope and the real relationships between the parts, so what it produces is grounded in your world rather than a plausible sounding average of everyone else's. It is also what keeps the things we design for you joined up. Your assessment, your study guide and your strategy all trace back to the same map and to one another, rather than being three separate documents that happen to share a title.

Yes, and it understands that they are not the same. The Australian side is anchored to ASQA and pulls live qualification and unit data from training.gov.au. The New Zealand side works to the NZQA framework. The regulatory graph holds both, so a provider operating across the Tasman is not stitching two different tools together.

No, and we would be wary of any tool that claimed it could. The expertise of your trainers and assessors is the thing that makes your delivery yours. VETos handles the drafting, the mapping and the formatting so your people spend their time on judgement, quality assurance and the learner, rather than on wrestling documents into shape.

Content comes out in formats your existing systems expect, including SCORM. The aim is to slot into how you already work rather than ask you to rebuild around us. If you have a specific platform in mind, raise it with us and we will tell you plainly what does and does not connect today.

Everything VETos generates exports to Microsoft Word (.docx) for everyday authoring and review, and to Adobe InDesign (.idml) for print-ready materials in high fidelity. The InDesign file hands straight to a designer to finish on-brand and to house style — with the layout and structure preserved, so there's nothing to rebuild by hand.

Your content and your data remain yours. We take security and data protection seriously, and we are happy to walk your IT or compliance team through exactly how data is handled, where it sits, and what protections are in place. If you are running a formal due diligence or security review, ask and we will provide the detail in the format you need.

You do. The training and assessment material you generate is yours, along with anything you bring in, such as your scope, your existing resources and your provider information. Supahuman owns the platform itself: the software, the regulatory graph, the way it is built and the improvements we make to it over time. Put simply, the tool is ours and what you make with it is yours. The exact terms are set out in our agreement, and we are happy to walk your legal or procurement team through them line by line.

Less time than you would expect for something doing this much. Onboarding is hands-on. Our team works with you to get your scope loaded, your settings right and your people comfortable, rather than handing you a login and wishing you luck.

VETos is priced on your scope, with volume discounts as your catalogue grows. You can see the rates and build an estimate on our pricing page. If you'd rather scope it together, get in touch and we will be straight with you.

VETos is built by Supahuman, an Auckland-based company building AI operating systems for regulated industries. We come from the VET sector's world, not from generic software, which is why the product is built around the standards, the scope and the audit, rather than bolting compliance on as an afterthought.

Book a demo. We would rather show you VETos working against a unit you actually deliver than talk at you about features. Bring a real scenario and we will run it.

Nova is the intelligent engine and delivery platform for next generation instructional design. It reads your source material the way a learning designer would, decides which concepts are worth turning into practice, builds the right experience for each one across audio, video, visual and guided interaction, and then delivers that experience live: the AI plays the counterpart in role plays, coaches beside every activity and narrates the video.

It is in production today with education and training providers, and it delivers straight to the learner: on a hosted link, as an app on their phone, or through your existing LMS.

None of the three. Authoring tools give you templates to fill; content generators hand you assets to package; an LMS hosts and tracks whatever you give it. Nova sits above all of them: it makes the design decisions, builds the experiences and delivers them live from a single link. There are no authoring tools in Nova, ever. If you want a change, you ask for it in a sentence and Nova rebuilds it.

It also plays nicely with what you already run: SCORM and LTI 1.3 are there for organisations that live in an LMS.

A frontier model is optimised to be broadly helpful. It can write a lesson and set a quiz, but it is not, on its own, an instructional designer, and it does not deliver a live, multimodal experience. Nova is the engineering that closes both gaps.

Nova reads for teachable concepts, matches each one to the right experience type (a judgement call wants a role play, a risky procedure wants a simulation, an explanation wants narrated video), scaffolds them into a learn, practise, show arc, grounds the design in the education standards of your market, and then delivers it: voice, coaching, consequences and debriefs, live. That is the difference between a generated asset and a delivered experience.

Nova is the capability; the number tracks the technology generation it stands on. Nova 5 is the current generation: the one that made high-fidelity, multimodal, live delivery possible.

We do not train foundation models. Nova sits above the frontier by design: it specialises the world's best models for learning with Supahuman's instructional-design engineering, so every time the frontier improves, Nova improves with it. Nova 6 and Nova 7 will follow, and everything built on Nova gets better without re-platforming.

eLearning has moved through four broad generations: static page-turners, multimedia courses, interactive click-through modules, and adaptive sequencing of the same. Nova 5 is built to define a fifth: contextual, emergent, culturally competent, live, tracked formative learning, generated for your industry and your learner, and delivered at a fraction of the cost and time of what came before.

In practice that means the learning is created for your world rather than assembled from templates, it responds live to what the learner says and does, and every experience is tracked back to the outcomes and competencies your framework requires.

Nova is a Supahuman product, built in Aotearoa New Zealand and Australia. Supahuman are experts in applied AI for education: our AI already serves education and training providers in a demanding, regulated sector, and Nova was engineered from day one to that sector's compliance standards, including NZQA and ASQA expectations, copyright, and data management.

Supahuman's legal name is Supahuman Limited, with teams in Auckland and Sydney.

Two worlds with the same problem. Corporate learning and development teams use Nova to turn onboarding, capability and compliance material into practice people actually do rather than click through. Education and training providers, including PTEs, RTOs, colleges and workforce trainers, use it to deliver the practise-by-doing learning their qualifications certify, mapped to their framework and defensible under review.

If you build learning for other people, Nova is for you.

Twenty activity types, chosen for the concept rather than the template, organised into a learn it, try it, show it arc.

Learn it: concept visuals, interactive models, animated explainers, video explainers, whiteboard lessons, lectures, process walkthroughs, explorable diagrams, study notes and podcasts. Try it: branching scenarios, hazard ID, sequencing, calculation practice, classify and sort, fault finders, document tasks, role plays and sim games. Show it: randomised quizzes with photo questions and the Quick Fire deck.

The common thread is that everything responds to the learner. The role play answers back, the sim's consequences compound, and a personal AI coach is one tap away on every activity.

The AI plays the counterpart by voice, in character, against clear session objectives: the difficult customer, the price objection, the distressed client, the toolbox talk. The learner speaks, the character answers back, and feedback is there the moment the learner asks for it.

Role plays are formative by design. They exist to grow capability through practice with feedback, judgement-free, not to grade people.

Yes, and it treats contextualisation as a first-class act. Narrated video and audio use your chosen voice and accent, so an Australian provider sounds Australian and a UK provider sounds British. Your branding sits on every learning space, and each space opens with a narrated welcome generated for your organisation.

Describe a client, team or cohort in plain language and every activity is re-set into their world, live in minutes, while the proven learning design underneath stays intact.

Yes. A learner uploads a test or a mock assessment and Nova returns structured formative feedback to their inbox within minutes: what is strong, what is missing, and what to do next, criterion by criterion against the unit's requirements. They can revise and upload again as often as they want.

It is formative only. Nova does not mark the work and never makes a competency decision: your assessors do that, exactly as they do today. The value is that the learner arrives at their real submission ready, which usually means fewer resubmissions and less repeated marking for trainers.

Because it runs on demand, the feedback is there at the hour people actually study, rather than waiting for the next time a trainer is free.

Ask for the change in a sentence and Nova rebuilds it. There are no authoring tools to learn and no timelines to book: revisions happen in plain language. New programme designs and revision requests get human sign-off before anything reaches a learner.

Nova reads your source material the way an instructional designer would and maps the concepts worth turning into practice, each with a rationale and a priority. It then chooses the right experience for each concept in the right modality and scaffolds the set into a coherent path, flexing for mixed ability, language and literacy, industry context, and the people around the learner.

You see the design before it is built: new programme designs get a human sign-off first.

Nova is formative by design. Practice comes with feedback that guides rather than grades, and progress grows a visible skyline, one bar per unit, rising as the learner practises. It is never presented as a mark.

For organisations that need evidence, every experience is mapped to the underlying outcomes and competencies, and the mapping is exportable. Coverage is evidence, not assertion.

Everything runs from a single link. There are no learner accounts and nothing to install to get started: open the link and the experience is live. Progress lives with the learner.

On a phone, a learning space installs as an app in two taps, with gentle practice reminders that know when to stop. No app store needed.

Yes. Deliver Nova learning spaces directly to your learners via web or mobile, or publish straight from your LMS: SCORM and LTI 1.3 are there for organisations that live in one. Nova sits inside the LMS where it must, and needs no accounts where it must not.

Bring the material behind one training module or one unit: the deck, the policy, the playbook or the unit standard. Nova turns it into a live learning space you can put in front of your team or real learners, typically the same day. That first space is the best way to evaluate Nova, because the proof is live product rather than a slide deck.

Yes. Nova generates companion guidance for the people around the learner, matched to context: parents and whānau for a school-age learner, workplace mentors and employers for an apprentice.

AI builds, AI reviews everything, and people decide anything the AI is not sure about. Every experience passes an AI quality review covering grounding, correctness, safety, pedagogy and cultural respect before it reaches a learner. Humans approve every new programme design, every revision request, and anything the reviewer is not certain about, and people set the standards the review checks against.

Design is pitched against the education standards of your market, such as the AQF, NZQF, RQF and their peers, and kept evergreen as standards change. Every experience is mapped to outcomes and competencies, and the mapping is exportable and ready for review. Coverage is evidence, not assertion.

Accessible as standard: transcripts on all narrated media, an accessibility options panel for learners, and reduced-motion support throughout. Experiences also flex for language and literacy as part of the design, not as an afterthought.

Cultural competency is engineered into generation, not bolted on. Nova works from a per-country cultural baseline, layers on each organisation's own cultural profile, and honours their Acknowledgement of Country word for word.

Nova is private by architecture. Learning runs from a link without accounts, progress lives with the learner, and there is nothing to install to get started. Regional expectations are respected per market, including the NZ Privacy Act, the Australian Privacy Principles and UK GDPR, with UK data residency available. For enterprise and institutional deployments, security review and data residency are part of the engagement.