amira01/meeting-assistant

Smart Meeting Assistant

This is a Gradio-based application that transcribes and analyzes meeting recordings or text transcripts. It supports both English and Arabic, using Whisper for audio transcription and a custom LLM for summarization and analysis.

Features

• —Audio Transcription: Upload an audio file to transcribe meetings using Whisper.
• —Text Analysis: Paste a transcript or analyze transcribed text to get summarized main points.
• —Multilingual Support: Processes content in English and Arabic with language-specific templates.

How to Use

1. 1.Audio Tab:
2. 2.Upload an audio file (e.g., MP3, WAV).
3. 3.Click "Process" to transcribe and analyze the content.
4. 4.View the transcript and analysis in the output boxes.
5. 5.Text Tab:
6. 6.Paste a meeting transcript.
7. 7.Click "Analyze" to get a summary with titled main points.

Setup Instructions

• —The application is containerized using Docker and deployed on Hugging Face Spaces.
• —Dependencies are listed in requirements.txt.
• —Ensure an OPENROUTER_API_KEY is set in the environment variables (via Space secrets).

Notes

• —Audio processing requires ffmpeg, included in the Docker image.
• —For Arabic inputs, the app ensures proper language detection and response formatting.
• —If deploying locally, set the OPENROUTER_API_KEY environment variable or modify the code to use a key file.

Built with ❤️ using Gradio and Hugging Face Spaces.