Speech Coach

AI-powered speech coach that refines your delivery, mimics your voice, and helps you become a confident public speaker.

Comment

Inspiration

Public speaking is challenging - many struggle with confidence and clarity.
We set out to create a tool that helps users refine their delivery with real-time feedback.

What it does

  • Speech setup - select the goal, tone and audience.
  • Upload and record - practice with slides.
  • Get feedback and a refined version in your own voice.

How we built it

The application is an MVP built using:

  • Python for backend processing.
  • Streamlit for a UI.
  • ElevenLabs for voice cloning and text-to-speech.
  • Fal for speech-to-text and text-based improvements.

Challenges we ran into

  • A lot of ideas and too little time - had to pick and choose.
  • Showcasing our product in demo video in concise manner.

Accomplishments that we're proud of

  • Coming up with the idea and turning it into POC within single day.
  • Complete pipeline working as intended with all of the features we planned in a current scope.
  • Used little API credits - our solution is relatively inexpensive.

What we learned

  • Working with ElevenLabs and Fal APIs which we haven't used before the competition.
  • Wide range of possible applications of AI Voice Agents.

What's next for Speech Coach

  • Expansion of the current pool of desired speech characteristics.
  • Analysis of the pronunciation and intonation - rigth now only speech content is being reviewed.
  • Analysis of the slides content and whether they match the speech.
  • Multiple language support.
  • Analysis of body language from a video.

Team Members

Built With

Share this project:

Updates