Evidence Pack and Reporting

Evidence Pack & Reporting

ASHU Mentor AI Studio separates candidate evidence capture from the digital presenter generation pipeline.

Evidence Categories

Evidence Type	Purpose
Candidate transcript	Supports evaluation and feedback generation
Candidate recording	Optional performance review and integrity evidence
Screen/focus events	Supports audit trail and candidate-session monitoring
Screenshots	Optional visual evidence for reports
Voice/video generation logs	Debugging and reproducibility for digital presenter generation
Final report	Summarizes candidate strengths, gaps, and recommendations

Generated Media Evidence

Every lecture render can store:

manifest.json
voice_generation_run.log
renderer_run.log
tts_audio.script.txt
tts_audio.wav
final MP4
browser-safe MP4

Example Manifest Fields

{
  "tts_audio": ".../tts_audio.wav",
  "authorized_voice_sample": ".../voice_samples/authorized_voice_recording.wav",
  "voice_generation_mode": "authorized_voice_sample_xtts_if_available",
  "renderer": "SadTalker",
  "browser_safe_video": ".../talking_presenter_output_browser.mp4"
}

Report Outputs

The candidate report should include:

1. Candidate profile summary
2. Interview round summary
3. Question-wise assessment
4. Coding-round observations
5. Communication and clarity feedback
6. Technical strengths
7. Improvement areas
8. Adaptive training recommendations
9. Evidence/log references
10. Final readiness rating