NextGen Learning: the complete guide to learning built into the work

What is NextGen Learning?

The classic model is familiar: a one-day training, a stack of slides, a test, a certificate. A week later most of it is gone. That is not a failure of the participants; that is how memory works. Information you hear once and never use fades fast.

NextGen Learning turns that around. It starts with what someone should do differently afterwards, divides the material into small pieces, has people retrieve and practise those pieces, and brings the learning back to where the work happens. It is not a new word for e-learning. It is a different design question: not "what do they need to know?" but "what do they need to be able to do, and how do we make sure it lasts?".

In CoVince these parts come together in one platform: a learning journey with swipeslides, AI help with creating, conversations with CrewMates, scenario games, 3D scenes, VR, rewards and reports. And because it is the same platform that holds documents, chat and workflows, learning is not next to the work but inside it.

The learning journey in six steps

A good learning journey follows six steps. You can click them on the main page and see which parts belong to each.

1. Set the learning goal. What should someone do differently afterwards? Phrase it as behaviour, not as a topic.
2. Design and create. From topic or document to learning content, with AI as the first designer and the subject expert as final editor.
3. Learn and practise. Learn in short bursts, practise safely, play together.
4. Apply at work. Learning in the place where the work happens, with workflows, chat and voice.
5. Measure and reward. Scores per theme, reports, points, awards and certificates.
6. Improve. Adjust based on what the figures and the people say.

The steps are a cycle, not a line. What you measure in step 5 becomes the input for step 1 of the next round.

Creating: from topic to learning journey with AI

The biggest obstacle to good learning content is usually time. The subject expert knows what needs to happen, but has no hours to pour that into slides and questions. In CoVince you give a topic or a document and the AI builds the structure, the chapters, the slides and the texts, with images and video added. You can translate a text with one click, have the AI adjust something per slide and have a preflight check go through the journey before you publish.

Important: the AI delivers the first draft, the expert decides. Existing material stays usable, because you can import a Moodle backup, a PDF or an Excel or CSV file with questions. Read more in AI-generated learning content without losing quality.

Learning and practising: four forms

Swipeslides are the basis: one idea per slide, 36 slide types, progress that is remembered. Microlearning with swipeslides explains why short pieces work.

CrewMates are AI characters you practise a conversation with, by typing or by voice, with assessment and a final report. See practising with an AI role-play.

Scenario games put a team in an incident, each with their own information, followed by a debrief. See scenario games for decisions.

3D and VR bring the place to the learner: a scanned space in the browser and VR, built in Unity and separate from the web platform. See 3D simulations with Gaussian splatting.

Applying: learning in the flow of work

Most training fails not in the room but on Monday morning. That is why learning belongs in the same environment as the work. A learning journey has a timeline, a chat and progress; workflows can assign slides to people; a Control Room can show learning KPIs. And from a voice report, the follow-up button takes you straight on to a swipeslide.

More on this in learning in the flow of work and transfer to the shop floor.

Measuring and rewarding

Measuring starts with a question: where does this group stand, and what do we need to change? Questions get a weight and a theme, open answers are assessed by AI, and for each learner the AI writes a personal report. For organisations with their own LMS there are SCORM, xAPI and LTI 1.3.

You reward with points, a leaderboard, awards with a reason, certificates and growth paths. The aim is recognition for what was learned, not only for speed. See measuring learning without invented numbers and rewards without the gamification pitfalls.

Why learning has to change now

Three developments make the old model untenable. First, work changes faster than an annual training programme can keep up with: what you learn today is partly outdated in a year. Second, attention is scarce: nobody has a day off for a training, but everyone has five minutes between two appointments. Third, technology can now do what only people could before: make a first draft of learning content, run a practice conversation and give direct feedback.

That last point matters, but it is a means. What remains is the question of whether people do their work better. NextGen Learning is the set of answers that comes closest to that: short, active, in the context of the work and with a measurement that says something.

What we know about how people learn

Good learning design rests on a handful of findings from learning psychology that have been solid for decades. You do not have to be an expert to use them.

• Retrieval strengthens. Whoever tries to recall something from memory remembers it better than whoever rereads it. That is why every slide is a question. See microlearning with swipeslides.
• Spacing works. The same practice time, spread over more moments, yields more than one block. See spaced repetition.
• Practice needs feedback. Deliberate practice has a concrete task, quick feedback and repetition. See feedback that works and practising with an AI role-play.
• Context counts. What you learn in a situation that resembles the work, you recall more easily there. See 3D simulations and transfer to the shop floor.
• Motivation comes from within. Rewards work as recognition, not as bribery. See rewards without the gamification pitfalls.

All the CoVince parts listed below are a way to put one or more of these insights into practice.

The parts at a glance

Part	What it does	Read more
Swipeslides	Short learning moments with one idea and a question, 36 slide types	Microlearning
AI generation	From topic or document to learning journey, with the expert as final editor	AI learning content
CrewMates	Practise conversations with an AI character, with a report	AI role-play
Scenario games	Deciding together with roles and a debrief	Scenario games
3D and VR	Learning in the place itself, in the browser and in VR	3D simulations
Rewards	Points, awards, certificates and growth paths	Rewards
Learning in the work	Timeline, chat, workflows and voice	Flow of work
Learning journeys	Route with branches and progress	Designing a learning journey
Sessions	Trainings, live games and teacher view	Classes and sessions
Measuring	Scores per theme, reports, SCORM, xAPI, LTI	Measuring learning

What it looks like per sector and situation

The parts stay the same, the emphasis differs. In onboarding it is about a route with an end, someone to ask and automatic assignment: see onboarding new employees. In compliance it is about short moments, decisions and evidence: see compliance training that sticks.

In healthcare difficult conversations, handovers and team situations are central: see NextGen Learning in healthcare. In education it is about time saved for teachers and a link with the LMS: see NextGen Learning in education. In industry and construction the place, safety and craftsmanship count: see NextGen Learning in industry and construction.

The people: trainer, manager and learner

Technology does not do the work alone. The trainer shifts from explaining to designing, coaching and final editing: see the role of the trainer. The manager is often the most important factor in whether something gets applied. And the learner needs control of their own path: choosing where and when, seeing where they stand, and getting recognition for growth.

For live moments there is the planning of trainings and sessions, with a teacher view: see planning classes and sessions.

Common mistakes

1. Starting with the technology. "We want to do something with VR" is not a learning goal. Start with the behaviour.
2. Building everything at once. A small learning journey that works is better than a big one nobody finishes.
3. AI without editing. Plausible is not the same as correct. The expert stays responsible.
4. Learning detached from the work. Without a connection to processes and managers it remains an event.
5. Measuring what is easy. The number of completed modules says little. Measure what fits the goal, and be honest about what you cannot measure.
6. Rewards without meaning. Points without explanation become noise.

A plan for the first ninety days

Days 1-30: choose and build. Pick one learning goal that matters and one group of ten to twenty people. Write the goal as behaviour. Have the AI make a first learning journey from your own material, and have a subject expert sharpen it. Add one form of practice that fits the goal.

Days 31-60: do and look. Let the group go through the learning journey, with a start and an end moment. Read the reports. What kept going wrong? Where did people drop off? Talk to a few participants and to their managers. Adjust the learning journey.

Days 61-90: apply and expand. Link the learning journey to the work process, for example through a workflow. Measure one behaviour at work. Decide what to expand: more groups, a second learning goal, an extra form. See designing the complete learning journey and measuring learning.

Honest about what is not (yet) there

A guide that only tells nice things is a brochure. So here are the limits. There is no ready-made library of 3D training scenes: you capture or upload each scene yourself. There is no separate timetable screen for classes: you plan with training dates, appointments and the calendar. The practice scenarios for conversations are a starting point of eight worked-out cases, not a full catalogue. The VR app is a separate application alongside the web platform.

And the AI is a tool with limitations. It writes fluently but not always correctly, it sees no face and feels no hesitation, and it assesses by the criteria you give it. That is why the expert stays final editor and a human coach stays valuable. Whoever knows this uses the technology better: for the repeatable, the fast and the scalable, and leaves the human to humans.

How to use this guide and the articles

Read this guide as a map and then go to the article that fits your question. Want to know how to build a learning journey? Start with designing the complete learning journey. Looking for a way to practise conversations? Read practising with an AI role-play. Need to back up a decision? Go to measuring learning. Each article stands on its own and points to related topics, so you choose your own route.

Want to see it in your own situation? Book a demo and bring a learning goal: in half an hour we will show what it looks like as a learning journey.

How do you start?

Start small and real. Pick one learning goal that matters, for example a conversation that often goes badly or a procedure where mistakes are expensive. Have the AI make a first learning journey, have the subject expert sharpen it, and let a small group go through it. See what the reports say and adjust.

You do not have to do everything at once. A learning journey with only swipeslides is already a step. Add a CrewMate conversation when it is about conversations, a scenario game when it is about decisions in a team. See designing the complete learning journey for the order.