Speech Improve

A project by William Nono, Ben Prystawski, and Ignas Panero Armoska.

For Hack the North 2018

Project Summary

This project uses IBM's Watson Tone Analyzer and Stanford's Empath project to create feedback on how a speech spoken might sound like to an audience, thus helping the speaker rehearse. The project uses the Google Cloud Platform and the Google Assistant to communicate with the user and is currently in a state where our python files are unhosted and so we are using ngrok for a webhook.

Accomplishments that we're proud of

We were able to come up with an idea, plan out its components and finally concretize it into a real product. But more importantly, we collaborated well, and ended up with a system that we can be proud of.

What we learned

  • We learned how to utilize Flask and IBM Watson's API
  • How to create webhooks for Google Cloud Platform's DialogFlow for the Google Assistant


In order to test it out and try out the system, open Google Assistant and say: "Hey Google, I'd like to talk to Speech Improve" and state something like : "I'd like to practice my speech" and insert or state speech.


Our team had never used Flask or the Google Cloud Platform before, and had limited experience with the DialogFlow System integration for the Google Assistant. We also had many roadblocks in terms of the communication between the Google Assistant and our Flask Application.

Future Possibilities Include:


If someone forgot a line, this project could be keeping track of where they were in a speech and remind them of the next part.

Timing Consideration:

If someone has a speech that must last a certain amount of time, this project could help them keep track of if they are speaking too long or too short.

Share this project: