See sample work Discuss your training need
Summary: Interactive workplace simulations for judgment-based skills. Learners make decisions, see consequences and receive a structured debrief through branching scenarios, role-plays or decision games delivered in SCORM-ready format.
Simulation Pilot

Let learners make the decision
before the consequence is real.

Turn an approved procedure, customer conversation or compliance dilemma into practice. Learners choose, see the operational consequence and receive a structured debrief before the real workplace moment.

✓ Branching decisions ✓ Role-play or decision game ✓ Consequences and debrief ✓ SCORM 1.2 or 2004 ✓ 15 business days ✓ Editable source files
Discuss a decision scenario See a sample
Choose the decision experience

Match the format to the behaviour that must change.

Tell us the workplace decision, the available options and what failure looks like. We recommend the format that produces the clearest practice and evidence.

🗺️
Narrative

Branching Scenario

A multi-scene interactive experience. Learners navigate through scenes, encounter domain characters, make decisions at branching points, read instruments, annotate images, and complete procedural tasks. Their choices affect state variables — Safety Score, Productivity, Compliance — and determine which path they take through the scenario.

Best for
Procedural knowledge, safety-critical decisions, compliance assessment
💬
Role-Play

AI Conversation

A live conversation with an AI persona. The AI character has a defined personality, scenario context, and communication rubrics. Learners are scored across rubric dimensions — Clarity, Empathy, Accuracy, Compliance Language — after the conversation ends. No scripts. Real practice.

Best for
Interpersonal skills, complaint handling, de-escalation, compliance dialogue
🎮
Game Engine

Decision Game

A round-based management simulation. Each round, learners make decisions — budget allocations, priority rankings, resource splits — fed through a formula engine that updates state variables. Conditional events fire based on thresholds. Final scorecard shows which decisions drove the outcome.

Best for
Management decisions, resource allocation, strategy under pressure
From procedure to decision practice

Your approved content defines the choices, consequences and scoring.

After a one-hour briefing call — or when you send us your source material (SOPs, procedures, PDFs, certification standards) — the generation pipeline runs. Twelve AI phases, each reviewed before the next starts. The result is a fully playable interactive scenario, QA-validated, compiled, and delivered as a SCORM package.

01
Knowledge Extraction
Identifies domain concepts, tasks, and learning objectives from your content.
02
Case Architect
Designs characters, backstory, scenario context, and the central conflict.
03
Fault Pattern Builder
Builds domain-specific defect or fault patterns for assessment scenes.
04
Scenario Graph
Designs the branching scene structure — which scenes lead where.
05–10
Storyboard → Interactions → Feedback
Writes dialogue, designs interactions, calibrates state variables, and writes feedback for every choice.
11–12
Compile & QA Validation
Builds the learner-safe SCORM manifest. Automated checks for dead ends, missing assets, unreachable pass scores.
What the learner experiences

A coherent workplace situation, not a string of quiz screens.

A Narrative simulation is not a single question type repeated. The generation system selects the right scene types for your domain and sequences them into a coherent scenario. This is what a learner can encounter in one simulation:

Branching Scenario
Multiple-choice decision point. Choice affects state variables and determines next scene.
Specimen Gallery
Gallery of specimen images. Learner identifies defects in each specimen.
Image Annotation
Learner draws zones on an image. Scored by Intersection-over-Union against correct zones.
Instrument Reading
Learner reads a simulated instrument (weld gauge, multimeter, etc.) and enters the reading.
Hotspot Explorer
Image with clickable hotspot regions. Learner must visit all required hotspots to proceed.
Procedural Stages
Multi-step task. Learner works through numbered stages in sequence.
Calculation Challenge
Learner enters numeric answers, marked correct within a tolerance % of the correct value.
Sequence Order
Drag-to-reorder task. Learner puts items in the correct procedural sequence.
Drag & Drop Task
Learner drags labelled items to matching target zones.
Character Dialogue
Character speaks to learner — informational scene with no scored interaction.
Consequence
Shows the result of a previous decision — immediate, contextual feedback.
Debrief
Closing scene: performance summary, path taken vs. optimal path.
Simulation · Permit to Work · Scene 3 of 9
Branching Scenario
The permit is missing Supervisor B's signature.
92
Safety
74
Compliance
📋
AI-generated scene image
Permit to Work document · Construction site
What do you do?
Stop work. Inform the site supervisor and wait for the permit to be correctly authorised before proceeding.
Proceed with the work — the other two supervisors signed and the task is time-sensitive.
Verbal authorisation from Supervisor B is sufficient to continue.

Sample: Permit to Work · Narrative simulation · Scene 3 of 9 · Safety Score state variable visible

What it actually is

The learner's choice changes what happens next.

Learners navigate scenes. At each decision point they choose. The state variables update — Safety Score drops if they skip a check, Compliance rises if they follow procedure. At the debrief, they see the path they took vs. the optimal path.

Characters

AI-generated domain characters — instructor, supervisor, worker — with personality descriptions that shape how they speak.

State variables

Learner decisions affect numerical KPIs throughout the scenario. The wrong call has visible consequences.

Assessment Mode

Hints and coach prompts hidden. Answer keys never reach the browser. Suitable for certification practice.

Branching flow map

Visual diagram of the complete scene graph. Every path is auditable before publication.

Industrial inspection certification

Virtual practical inspection — welding, rigging, scaffold

Give candidates repeatable practice with visual defects, inspection sequences and evidence-based scoring before they enter a physical assessment. The simulation supports preparation; it does not replace jurisdiction-specific practical certification requirements.

Every inspection competency is covered by the existing scene type library. Candidates identify defects from specimen images, annotate defect locations on photographs, read simulated instruments, classify severity, and make accept/reject decisions — each scored independently against the relevant standard.

Defect recognition Specimen Gallery
Defect location Image Annotation
Instrument reading Gauge simulation
Severity classification Decision branching
Accept / Reject Scored decision
Procedure sequence Step ordering
Safety awareness Hotspot Explorer

Welding Inspection

AWS CWI · CSWIP 3.1 · PCN WT Level 2

20 defect types: porosity, undercut, overlap, lack of fusion, cracks. Specimen gallery, weld gauge reading, accept/reject decision.

Rigging & Lifting

LEEA · ASME B30 · CPCS

12 defect types. Wire rope condition, sling rating verification, load angle calculations, documentation accuracy.

Scaffold Inspection

CISRS · NASC · IPAF

12 defect types. Visual hazard identification, component condition assessment, guardrail compliance.

Simulation Pilot
One complete decision simulation

One complete simulation — your choice of Narrative (branching scenario), Role-Play (AI conversation), or Game Engine (decision game). Delivered in 15 business days.

  • ✓ Choice of simulation type: Narrative, Role-Play, or Game Engine
  • ✓ AI-generated from your source content or briefing
  • ✓ Narrative: up to 15 scenes with characters, instruments, defect specimens, hotspots
  • ✓ Game Engine: formula-driven state variables, conditional events, round-based decisions
  • ✓ Role-Play: AI persona with rubric-based conversation scoring
  • ✓ Assessment Mode — hints hidden, answer keys server-side only
  • ✓ SCORM 1.2 or SCORM 2004 (Narrative and Game Engine)
  • ✓ Hosted learner link — no LMS required
  • ✓ All source files — you own everything
  • ✓ Delivered in 15 business days
  • ✓ Optional localisation scoped by language and reviewer requirements

Want to pair it with teaching content? Add a Course Pilot as a separately scoped production track.

Questions

Frequently asked

What is a Simulation — is it a video?+
No. A simulation is an AI-generated interactive scenario that runs in the browser (or in any LMS via SCORM). Learners make real decisions at branching points, annotate images, read simulated instruments, work through procedural tasks, and complete drag-and-drop exercises. It is a structured interactive experience with AI-generated images, characters, and scenario content — not a passive video. We do produce a demo recording of a simulation being played, but the deliverable to your learners is the interactive SCORM package.
What's the difference between the three simulation types?+
Narrative (Branching Scenario) is a multi-scene scenario with characters, images, instruments, and decisions that affect state variables — best for procedural and safety-critical skills. Role-Play is a live AI conversation scored on communication rubrics — best for interpersonal and compliance dialogue skills. Game Engine is a round-based decision simulation with formula-driven outcomes — best for management and resource allocation decisions. You specify which type when you brief us, and we generate the right one for your learning objective.
What do you need from me to generate a simulation?+
A one-hour briefing call. If you have source material — SOPs, inspection procedures, policy documents, PDF standards — send them and the simulation will be more domain-accurate. If you have nothing, we build from the briefing alone. For industrial inspection simulations, we can also work directly from the relevant certification standard.
How does the AI build the simulation from my source content?+
The generation runs 12 sequential AI phases: knowledge extraction from your content, case design, fault pattern building, scene graph design, storyboarding, asset planning, image generation, interaction design, consequence calibration, feedback writing, manifest compilation, and QA validation. The whole process takes 4–12 minutes of compute time. We review the output before delivering to you.
What scene types are available in a Narrative simulation?+
Fifteen scene types: branching scenario, case intro, investigation, consequence, hotspot explorer, specimen gallery, image annotation, procedural stages, calculation challenge, instrument reading, sequence order, character dialogue, drag-and-drop task, timer challenge, and debrief. The generation system selects scene types appropriate to your domain — you don't need to specify them.
Can you build an inspection simulation for industrial certification — welding, rigging, scaffold?+
Yes. This is a specific capability. We have a defect catalog covering 44 defect types across welding (20), rigging (12), and scaffold (12), mapped to AWS CWI, CSWIP 3.1, LEEA, CISRS, and other certification standards. The simulation includes specimen galleries, weld gauge readings, image annotation for defect location, severity classification decisions, and accept/reject decisions. Each defect maps to per-parameter competency scores, not just a total percentage.
Can I load the simulation into my own LMS?+
Yes. All Narrative and Game Engine simulations export as SCORM 1.2 or SCORM 2004 packages. The package runs entirely within your LMS — no connection back to Maventra needed during learner play. Completion and score data report back to your LMS via the SCORM API. Role-Play simulations deliver as a hosted learner link.
What is Assessment Mode?+
Assessment Mode disables hints, coach prompts, and worked examples so learners receive no guidance during the simulation. It turns the simulation into a genuine assessment instrument. The answer keys and scoring logic are always server-side — never sent to the browser — so Assessment Mode cannot be circumvented.
Why does the Simulation Pilot take 15 days instead of 10?+
The simulation generation pipeline runs 12 AI phases, each reviewed before the next starts. After generation, there is a QA validation pass, an SME review window, and a compile step before the SCORM package is ready. The extra 5 days reflect that review cycle — not a quality trade-off. Course Pilots and Certification Prep Packs both deliver in 10 business days.
Start with the decision

What must the learner decide correctly when the pressure is real?

Share the procedure, role and consequence. We will map the scenario, decision points, evidence and best delivery format before production begins.

Discuss a decision scenario See our delivery standard