We have noticed multiple instances of cheating by participants, so the final stage of evaluating the stories will be done manually.
How manual judging was done:
- 5 judges read and evaluated 3-7 stories from each team.
- Stories were chosen randomly and independently for each judge.
- Each judge scored quality and engagement of stories from 1 to 10.
- 3 judges evaluated the tech part of solutions from 0 to 2
- it doesn't work or data is hard-coded.
- it works but there is a simple prompt and just API calls to existing AI model.
- it works and there is an innovative idea
- Final score = Average score for quality multiplied by tech score.
Final rating:
[team is nickname on Story3]