How sujanprasad/vbcua works
VBCUA is a local speech-assessment dashboard. A candidate selects a reference concept, uploads an audio response, and runs analysis: the app transcribes speech, compares its meaning with the reference, extracts pacing and volume features, calculates a score and feedback, and saves the assessment. The app then presents diagnostics and offers a PDF report; users can also manage concepts and review or delete assessment history. The sampled frontend and analysis modules support this map; the README documents the end-to-end workflow and capabilities. Some internal call details are not shown, so those links cite the README.
Main entry points: app.py Streamlit dashboard · transcription.py Speech transcription · semantic_analysis.py Semantic analysis · audio_features.py Audio features
Components 8
Assessment interface
- Streamlit dashboard
5.Project_Development_Phase/frontend/app.py
Analysis pipeline
- Speech transcription
5.Project_Development_Phase/backend/modules/transcription.py - Semantic analysis
5.Project_Development_Phase/backend/modules/semantic_analysis.py - Audio features
5.Project_Development_Phase/backend/modules/audio_features.py - Scoring and feedback
5.Project_Development_Phase/backend/modules/evaluation.py
Assessment data
- Relational records
5.Project_Development_Phase/backend/database.py
Reporting
- PDF report generation
5.Project_Development_Phase/backend/modules/report_generator.py
Other components
- Candidate
Connections 8
- Candidate → Streamlit dashboard: uploads response
README.md - Streamlit dashboard → Speech transcription: requests transcription
README.md - Streamlit dashboard → Semantic analysis: requests comparison
README.md - Streamlit dashboard → Audio features: requests extraction
README.md - Streamlit dashboard → Scoring and feedback: requests scoring
README.md - Streamlit dashboard → Relational records: reads and writes
5.Project_Development_Phase/frontend/app.py - Streamlit dashboard → PDF report generation: generates report
README.md - Streamlit dashboard → Candidate: shows results
README.md