O
OMNILEARN3 · HACKATHON SHOWCASE

From lecture content to useful learning.

A local, example-led assessment that helps turn learner signals into a better brief for teaching materials.

No API keyLocal browser appHuman-guided
01 · The challenge

One lecture.
Many starting points.

Students arrive with different backgrounds, confidence levels and ways of making sense of difficult material.

01Surface fluencyA correct answer may hide fragile reasoning.
02Confidence gapsSome learners know more than they feel able to explain.
03Transfer mattersRecall is not the same as using an idea in a new context.
02 · What OmniLearn3 does

It asks for examples, not labels.

The learner responds to scenarios, trade-offs, rankings, error diagnoses and confidence checks. The app quietly infers useful instructional signals.

INPUTExamplesShort decisions and miniature cases
→
INFERENCELearner briefFamiliarity · challenge · support needs
→
OUTPUTBetter promptSpecific instructions for the next materials agent
03 · The experience

Compact questions.
Useful evidence.

12Signal itemsScenario judgements, rankings, worked examples and transfer checks.
4Separate signalsPrior familiarity, challenge, instructional format and readiness.
1Reusable handoffA copy-ready prompt for study materials and assessment design.
Back and reviseConfidence ≠ correctnessLocal storage
04 · The output

A learner brief that makes the next step clearer.

“Use more worked examples, explicit scaffolding and retrieval practice, then introduce transfer once the core mechanism is secure.”

One prompt can ask the next agent to create:

Study notes · glossary · misconceptions · practice questions · answer key · revision guide · instructor assessment · active-recall flashcards

Mapped to learning outcomesCited to source material
05 · Major lesson learned

AI does not replace teaching expertise.

The useful boundary

  • Agents can accelerate analysis, drafting and personalisation.
  • They cannot decide what matters educationally without subject and teaching judgement.
  • Evidence, context, feedback and human review remain essential.

The opportunity: use agents to multiply thoughtful teaching — not to remove it.

06 · Closing thought

With guidance, personalisation becomes practical.

OmniLearn3 turns student input into a clearer brief for the next learning-materials workflow.

Reveal useful signalsKeep humans in the loopBuild better materials

A prototype lesson: the strongest educational AI is collaborative, evidence-aware and guided by expertise.