Main Prize Criteria
Teams must demonstrate the following in their pitches:
Convincing framing of problem (20%)
Teams need to demonstrate that the problem is specific and well-defined, with evidence or research supporting its importance.
- Is the specific social issue and pain point clearly defined?
- What evidence or research demonstrates this problem’s significance and scope?
- Who are the stakeholders affected by this problem, and how?
- What existing solutions have been attempted and why have they been ineffective?
- Does the project address the root causes rather than just symptoms?
- Why does this problem deserve attention now?
Prototype/ demo (20%)
Team must demonstrate a functional prototype or demo illustrating the proposed solution in action.
- Does the demo effectively showcase the solution’s core functionality?
- How complete is the prototype (what works vs. what’s simulated)?
- Can the prototype handle realistic test cases?
- How intuitive is the user interface/experience?
- Has the prototype been tested with potential users?
Note: Teams without a technical prototype will be heavily penalised in both the Prototype/ demo and Appropriate technical component criteria.
Appropriate technical component (20%)
Teams must show that GenAI technology is effectively and appropriately used for the problem.
- How specifically does the solution leverage GenAI capabilities?
- Why is GenAI particularly suited for addressing this problem?
- Which specific GenAI models or techniques are being implemented?
- How does this approach improve upon non-AI alternatives?
- What technical limitations exist, and how have they been addressed?
- How efficiently does the implementation use computational resources?
Feasibility and impact (20%)
Teams must convincing argue that the solution is realistic to implement and likely to make a meaningful social impact.
- What is the business model or plan?
- Is the solution practically implementable given real-world constraints?
- How will target users access and adopt the solution?
- What metrics will measure impact?
- What is the potential scale of impact (number of beneficiaries)?
- What implementation challenges are anticipated?
- Has the team developed a realistic timeline for deployment?
Sustainability and ethics (20%)
Teams must show the solution is designed for long-term sustainability and state its ethical implications.
- Does the business model or plan support this solution long-term?
- Is the solution scalable, adaptable, or has potential for integration with existing systems or processes?
- How will this project be financially sustained?
- What potential ethical concerns (e.g. bias, privacy, and accessibility, unintended consequences) exist with this GenAI implementation?
- How has feedback from affected communities been incorporated?
- Does the solution empower users rather than creating dependencies?
WorkBuddy Special Prize
Teams must demonstrate the following in their use of the WorkBuddy technology:
- Purposeful Human-AI Collaboration (30%) – Use for right tasks and to provide useful context but maintained human judgement.
- Reusable Skills, Experts, Agents or Automation (25%) – Create practical and repeatable assets that others can use in the future.
- Quality of Outcomes (20%) – Demonstrate impact to team’s research, execution, prototype, collaboration, or final pitch.
- Responsible Use and Validation (15%) – Verify important outputs, handle information appropriately, and recognise risks and limitations.
- Precision and Efficiency (10%) – Use of clear prompts, structured context, and purposeful iteration (instead of unnecessary repetition).
Guidance:
- Teams will be assessed on how they use WorkBuddy to solve real project challenges, validate important outputs, improve their work and create reusable Skills, Experts, Agents, automations or workflows for others.
- Quality over quantity: More prompts, features or tokens do not automatically lead to a higher score.
- Reuse matters: Judges will give particular credit to a useful method that other people can adopt.
- Evidence matters: Screenshots, workflow records and before-and-after examples are stronger than general claims.
- Human judgement remains essential: Teams are responsible for the accuracy, safety and appropriateness of their work.
The WorkBuddy Award does not affect the judging for the main prizes.