Production
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Uplift Modeling | Production | LIFT@300.0286 | 12 | |
| CTR Prediction | Production-5k | GAUC0.6377 | 6 | |
| WorkbackPlan | Production WorkbackPlan (test) | F1 Score79.8 | 4 | |
| StatusReport | Production StatusReport (test) | F1 Score79.4 | 4 | |
| Production Email (test) | F1 Score79.8 | 4 | ||
| DirectReports | Production DirectReports (test) | F1 Score79.5 | 4 | |
| UserManager | Production UserManager (test) | F1 Score79.7 | 4 | |
| Calendar | Production Calendar (test) | F1 Score79.4 | 4 | |
| WebSearch | Production WebSearch (test) | F1 Score79.4 | 4 | |
| Answering Machine Detection | Production (val) | Accuracy97.6 | 4 | |
| Text-to-SQL | Production | Mean Score91.3 | 3 | |
| Sign alignment with online outcomes | Production A/B experiment (904 queries) | Sign Alignment0.368 | 2 | |
| Document Processing | Production up to 21 January 2026 (deployment) | Count926 | 2 |
Showing 13 of 13 rows