- Credit everything. Open-source libraries, APIs, templates, assets and datasets must be credited in your README. Any pre-trained model or dataset must be publicly available and disclosed.
- AI tools are allowed and expected — Copilot, Claude, ChatGPT, Cursor etc. Two conditions: disclose it in your README, and be able to explain your own code. Judges may ask how something works, and "the AI wrote it" isn't an answer.
- Commit as you go.
- Late submissions aren't considered unless you reported a technical issue to organisers before the deadline. Submit a draft early — Devpost lets you edit until 11:59pm.
- Plagiarism or misrepresentation means disqualification. All participants follow the WIT Code of Conduct.
Stage 1 — Shortlisting (Monday). A WIT panel reviews every submission and picks the finalists.
| Criterion | Weight |
|---|---|
| Innovation & Creativity | 20% |
| Technical Execution | 20% |
| Impact & Relevance (against your track) | 20% |
| Design & UX | 15% |
| Presentation & Communication | 15% |
| Iteration & Feedback | 10% |
Stage 2 — Finalist pitches (Tuesday). Finalists pitch live to industry judges: 5 minutes + 2 minutes Q&A, scored on what happens in the room.
| Criterion | Weight |
|---|---|
| The Demo | 30% |
| Problem & Idea | 25% |
| Pitch Craft | 20% |
| Design & Experience | 15% |
| Depth in Q&A | 10% |
Your pitch must include a demo. Live is best — but recorded demos are not penalised.
Who can enter- Teams of 2–4. No solos.
- All members must be current WIT members — membership is free, sign up here.
- Teams must be at least 50% female or non-binary.
- Cross-disciplinary teams are strongly encouraged. Designers, writers, business students — you make projects better, and the judging rubric rewards it.
- To be eligible for prizes your team must be have at least one member in Melbourne.
- One team per person.