Public signal scores
Current public signal rankings by category. Movement compares the two latest completed score calculations; every history graph uses the same 0–100 scale.
Category rankings
Companies are ordered from the highest current score to the lowest. Select a company to review only its public signal history.
Category
Companies that record how people interact with software, including screens, clicks, keystrokes and application sequences, to teach AI agents how to execute the same tasks.
| Company | Score | Movement | History |
|---|---|---|---|
| 78 | — | — | |
| 68 | — | — | |
| 67 | — | — | |
| 66 | — | — | |
| 56 | — | — | |
| 48 | — | — |
Category
Companies that extract the reasoning, judgment, rules and undocumented know-how behind human work, usually through interviews, conversations or feedback rather than direct observation of software actions.
| Company | Score | Movement | History |
|---|---|---|---|
| 68 | — | — | |
| 63 | — | — | |
| 52 | — | — | |
| 46 | — | — | |
| 32 | — | — |
Category
Companies that recruit skilled humans to create original data for AI training or evaluation. The work is produced specifically for model development rather than captured from existing workflows.
| Company | Score | Movement | History |
|---|---|---|---|
| 88 | — | — | |
| 78 | — | — | |
| 78 | — | — |
