The goal of this app is to help people practice their speech and presentation skills by introducing AI Agent that provides feedback according to pre-defined settings.

The app gathers a sample of your speech, refines it, and reads it in your own voice.

In order to run the app please provide your own API keys to ElevenLabs and FAL modules in .env and run source activate_env.sh . The environment was built on Ubuntu.

Execute streamlit run app.py and proceed with the in-app instructions. The app requires you to:

• choose the context of the speech out of the predefined tone and audience types
• upload a PDF file to help you perform the speech - it is not used in any other way than just for display.
• upload a speech sample that is gathered in-app

The app offers:

• text feedback based on the desired characteristics that were specified on the first page
• a voice sample of refined speech in user's own voice.

Features that will be added in the future:

• Expansion of the current pool of customizable settings
• Analysis of pronunciation of the speech
• Analysis of the content of the PDF slides and whether they fit the speech
• Ability to have a conversation with the agents that will further refine the output speech
• Multiple language support
• Analysis of body language from a video