Purpose
Use this hackathon judging rubric to make demo expectations clear before teams start building. It helps participants understand what matters, gives judges a shared language, and keeps feedback focused on useful execution instead of presentation polish alone.
This rubric is designed for community hackathons, AI builder events, and practical Codex workflow demos where teams need a fair, public-safe way to explain what they built.
Criteria
- Usefulness: the project solves a clear problem for a real user or workflow.
- Clarity: the demo and story are easy to follow in a short presentation.
- Execution: something real works, even if it is small or rough.
- Learning: the team can explain tradeoffs, constraints, and what they would improve next.
- Responsibility: the project avoids unsafe behavior, private data, and misleading claims.
Suggested scoring
- Usefulness: 1 to 5 points.
- Clarity: 1 to 5 points.
- Execution: 1 to 5 points.
- Learning: 1 to 5 points.
- Responsibility: pass, revise, or disqualify if the project exposes private data or unsafe behavior.
Keep the scoring simple enough that judges can apply it during live demos without a long private deliberation process.
Judge prompts
- Who is the user and what changed for them?
- What part of the demo is working now?
- Where did Codex or another AI tool help the team move faster?
- What tradeoff did the team make intentionally?
- What would be the most useful next improvement?
Participant guidance
- Show the working artifact first.
- Explain the user problem in one or two sentences.
- Name the workflow, tool, or data source that mattered most.
- Be honest about what is mocked, manual, or incomplete.
- Avoid private customer data, partner details, credentials, and internal plans.
Public-safe notes
This rubric should stay public and reusable. Do not add private judge names, private scoring sheets, sponsor-specific preferences, attendee data, or internal judging notes to this file.