How shadlia/transcription_chat_llm works
Project type and purpose
• This is a full-stack web application called “transcription_chat.”
• On the frontend it’s a React app (AudioRecorder, AudioUpload, Chat, TranscriptDisplay, etc.).
• On the backend it’s a Python (likely Flask or FastAPI) service handling transcription, chat and summary routes.
• The app lets a user record or upload an audio clip, transcribe it, chat about it (ask questions), and generate summaries.
• It’s ideal when you want to model user goals (record audio, view transcript, ask questions, get summaries) rather than class/module details.
Main Actor(s)
• End User (the only external actor)
• (Optionally) System Administrator or Developer for maintenance tasks, but focus on End User here.
Use Cases
Them by functional area:
A. Audio Input
• Record Audio
• Upload Audio File
B. Transcription
• Request Transcription
• View Transcript
C. Chat Interaction
• Start Chat Session
• Ask Question about Transcript
• View Chat Response
D. Summarization
• Generate Summary
• View Summary
E. Parameter Management
• Adjust Transcription/Chat Parameters (e.g., temperature, model)
F. Navigation
• Navigate to Home
• Navigate between pages
Relationships
• Use «include» relationships:
• “View Transcript” «includes» “Request Transcription”
• “View Chat Response” «includes» “Ask Question”
• Use «extend» relationships if a use case is optional or conditional:
• “Generate Summary” «extends» “View Transcript” (only if the user chooses summary)