Innovation Lab · Experiment 001
Prototype · Preview
When software changes, what should learning change?
A button gets renamed. An approval step appears. Most of the time, the course doesn't need a rebuild — but sometimes it really does. Pick a fictional workflow change, see its instructional risk, and record a human decision.
Rule-based prototype. Not a live AI agent, and not a Storyline file editor.
Baseline task
Analyze a change to see what moves
The workflow learners practice today
1
Select a software change
2
Instructional impact
No analysis yet
Choose a change and run the analysis. Nothing is updated or published automatically.
- Objective
- Proposed change
- Validation
Human approval required. Classification is illustrative and rule-based, not AI-generated.
3
Human review
Review decision
Review logEmpty
Recorded reviews appear here. They stay in this browser session until exported — nothing is sent to a server or published to Storyline.
4 · What we will measure
The question isn't whether it's fast. It's whether it's right.
- Critical-change recallDid high-risk changes get flagged?
- False alarmsHow often low-risk changes triggered rebuilds.
- Reviewer correction rateHow often humans overruled the classification.
- Time to approved planFrom detected change to signed-off update.
- Decision alignmentWhether practice still targets the real decision.
This prototype demonstrates a workflow, not measured effectiveness.