The inspiration behind Verbate stems from personal journeys of our team as students. Anton's story: Public speaking was a skill I struggled with for a long time. Over the years, I took numerous lessons and attended workshops to improve, but progress was slow. What held me back was the lack of opportunities to practice on my own, efficiently and without incurring additional costs. If I had a tool like Verbate back then, I could have honed my public speaking skills more effectively. All of us in the team have had similar experiences or know somebody who did, which drove us to create Verbate, a resource that empowers students, like Anton, to practice speaking confidently and eloquently without any cost and time challenge. With Verbate, students can gain valuable experience, receive constructive feedback, and become better communicators on their own terms, all for free.

🤔What it does🤔

Verbate is a web app that helps users improve their interview and public speaking skills by simulating real verbal conversations. It uses a microphone to listen and text-to-speech to give feedback and personalized guidance for more effective communication. Verbate allows you to select different interviewing styles, to select the difficulty and how harsh the judge is.

🔨How we built it🔨

  • Built with node.js, next.js, index.js, React, Radix UI
  • JavaScript for Backend,
  • HTML, CSS, for Frontend
  • Used GPT-4 (OpenAI API Key), Amazon AWS Polly (API Key)

💪Challenges we ran into💪

  • Difficulties in fine-tuning the AI feedback system to provide nuanced and helpful responses.
  • Microphone wouldn't stop recording even after the user is done responding
  • On top of everything, making the web-app user-friendly and accessible required continuous refinement.
    • Library dependency issues
    • Display bugs on the web app
    • Uploading the app to an actual website was very slow

🏆Accomplishments that we're proud of🏆

Naturally, we are proud that our app works fully, and by extension, to have successfully harnessed the power of AI in order to develop this tool that allows users to take charge of building their own confidence in speaking. Witnessing the app in action, listening to the user and giving direct feedback with a voice fills us with a sense of fulfillment, knowing that we have completed what we sought out to do, and all in only 24 hours.

One of our proudest achievements is the ability to offer Verbate to students all over the world for free. We're committed to breaking down financial barriers to education and ensuring that students, regardless of their background or resources, can access a Verbate to enhance their interview and public speaking skills. Our mission is to empower students globally and to provide them with the support they need to succeed, in both academic and professional endeavors.

Although Verbate is primarily used to give feedback on effective communication and interviews, the way we built this project allows it to be extended to other topics very easily. We are proud of the flexibility of Verbate

🎓What we learned🎓

While creating this project, we learned many things:

  • First and foremost, we learned a lot about AI technologies, particularly in natural language models and processing, and how to implement them in programs (in this case, to analyze answers and provide feedback).
  • Second, as with any group project, we had different ideas and disagreements, which came from the very start at brainstorming, and kept coming throughout the entire project. We disagreed on its name, its logo, and the necessity of many parts of the app. However, in order to work optimally, we needed teamwork and chemistry, so this experience served as a reminder that by communicating effectively and planning things with the team before doing them, the development process could have been much faster, leaving us with no time crunch. Often times, we acted as if we were working on three individual projects despite being a team.
  • Lastly, we learned the value of a user-centered approach. Since we worked on the backend for most of the time, we had a very bad UI that would have been unnavigable, but luckily, we caught ourselves. Then, we tried to start from the user's experience and work backward to give users the best experience. Although this made some of the original program buggy, we fixed everything and this proved to be a lot more effective in creating a better overall feel to the final product than trying to engineer the idea before adding emphasis on UI features.

🔜What's next for Verbate?🔜

There is always room for improvement, and as stated before, the way we built this project allows for it to be extended to more than just Interview practice and feedback. In the future, we would like this to become an app that can be downloaded on all platforms, to make it as accessible as possible. Future development ideas include:

  • More advanced feedback: Enhancing the AI feedback system with more advanced natural language processing and analysis for sentiment and tone in the voice to provide even more precise and constructive suggestions, particularly helping people with the tone of their voice.
  • Feedback analytics & Account system: Providing users who have logged in with analytics on their progress and improvements from previous sessions.
  • Integration with video and other AI tools to make for an even more authentic and seamless interview experience.
  • Multilingual support
  • Feedback feature from users

By taking these next steps, we can improve the experience of practicing interviews and public speaking, reach more individuals in need, and making a lasting positive impact on students and professionals all over the world.

Overall, the development of Verbate was a fantastic learning experience for us, with a very valuable outcome for us, and hopefully for the world! Thank you for reading through this project's description. Try it out yourself with the files linked in github below!

Built With

Share this project: