What it does
A hand-written keyword classifier faces the same 48 tickets. In the maker's comparison, Jev routed 45 correctly and the baseline routed 40.
Playable demos & bots
A browser colony simulation where support tickets carry real text and Jev assigns the categories agents use to route them.
Only you can see your notes.
Screenshot unavailable. Open the experiment ↗
A hand-written keyword classifier faces the same 48 tickets. In the maker's comparison, Jev routed 45 correctly and the baseline routed 40.
Borrow this approach to compare two ways of sorting your project's messages. Start with the demo's sample tickets and hidden answer key. A developer can adapt the simulation so each sorter sees the same messages but never the correct answers. Watch how wrong guesses send workers to the wrong place.
Keep your answer key separate from the guesses when trying new examples. The demo compares simple word-matching rules with AI. A higher score on these few tickets does not show that the AI will sort your messages correctly.