Week 4 · Student Packet

Exhibition Day
Final Checklist & Score Sheet

Part 1: Your pre-pitch checklist. Part 2: Score sheet for pitches you watch. Part 3: Reflection.

Team Name: _______________________
Date: _______________________
Part 1 — Pre-Pitch Final Checklist

Complete this before you go on. Every box should be checked.

Content
Problem is specific and opens with a hookFirst slide or first sentence makes the audience feel the problem — not just understand it.
ML type is named and justifiedNot just "AI" — the specific type of ML and why it's the right approach.
Training data sources are namedSpecific datasets or collection methods — not just "we'll find data online."
Bias type is named with a specific mitigationHistorical, sampling, label, feedback loop, or measurement bias — and a real strategy to address it.
Ethics: at least one harm scenario and one safeguardWho could be harmed if the system fails? What's the safeguard?
Two specific failure cases describedConcrete scenarios — not just "sometimes it might be wrong."
Future work is ambitious and specificNot just "add more features" — what, why, and what it would require.
LLM-specific concerns addressed (if applicable)Hallucinations, cultural bias, prompt injection, or environmental cost — whichever apply.
Slides
Every slide has a visualDiagram, mockup, chart, or user flow — not just text.
No slide is a wall of bullet pointsIf a slide has 6+ bullets, it needs to be redesigned.
Font is readable from 3 metres awayBody text minimum 20pt. Headers 28pt+. Check from the back of the room.
Deck is loaded and tested on the display deviceDon't find out there's a font issue mid-pitch.
Delivery
Timed at 5–7 minutes on a practice runNot "probably around 6 minutes." Actually timed.
Every team member knows their sectionJudges will direct Q&A to quiet members.
Opening line memorisedThe first 15 seconds set the entire tone. Know it cold.
Q&A plan readyLikely questions: "How would you handle X bias?" "What if your system is wrong for group Y?" Who answers these?
Part 2 — Audience Score Sheet

Score each team as they pitch. 3 = Strong, 2 = Developing, 1 = Weak.

You'll use this to inform your audience vote at the end. Be honest — your vote counts 50%.

Team Score Sheet #1
Team name:
Problem Clarity
Real, specific, compelling
3
2
1
AI Solution + ML Type
Named, justified, explained
3
2
1
Data & Bias
Sources + type + mitigation
3
2
1
Ethics & Impact
Harm named + safeguards
3
2
1
Limitations & Failures
2+ specific failure cases
3
2
1
Future Work
Ambitious and specific
3
2
1
Pitch Quality
Clear, confident, visuals
3
2
1
Your Total (max 21)—
Team Score Sheet #2
Team name:
Problem Clarity
3
2
1
AI Solution + ML Type
3
2
1
Data & Bias
3
2
1
Ethics & Impact
3
2
1
Limitations & Failures
3
2
1
Future Work
3
2
1
Pitch Quality
3
2
1
Your Total (max 21)—
Team Score Sheet #3
Team name:
Problem Clarity
3
2
1
AI Solution + ML Type
3
2
1
Data & Bias
3
2
1
Ethics & Impact
3
2
1
Limitations & Failures
3
2
1
Future Work
3
2
1
Pitch Quality
3
2
1
Your Total (max 21)—
Part 3 — Post-Exhibition Reflection

Complete this after the results are announced.

AI Innovation Challenge · Week 4 of 4 Thank you for building something real. 🏆