AI Decision Telemetry & Evaluation
Real-time decision distribution, human review gating efficiency, and accuracy metrics.
Total Triaged
0
Total persisted decisions
Safety Review Gate
0.0%
0 issues flagged for review
Human Agreement
100.0%
0 human reviews verified
AI Provider Mode
Dual-Mode
Jev 1.13 / Local Heuristic
Priority Distribution
No priority data recorded yet.
Category Classification Distribution
No category data recorded yet.
Evaluation Benchmark Suite
IssuePilot evaluates decision models against a curated ground-truth dataset (data/evaluation-benchmark.json).
$ npm run evaluate
Running 15 ground-truth benchmark cases...
Category Accuracy: 60.0% • Escalation Recall: 100.0% • Avg Latency: 0.1ms