The idea for PerfectSpeech AI stems from our personal experiences. As the team lead, I (Lynnshelsie) have struggled with a speech impediment, and Jaeden, our team member, is an international student who has insecurities about his accent. Both of us face challenges with public speaking, which is something many students can relate to. Being in college, we understand the critical importance of clear communication, especially in academic and professional settings. Our goal was to create a solution that would support others facing similar challenges.
We wanted to design a tool that could help people transcribe their spoken words easily and accurately, making communication more accessible for everyone, no matter their background. By focusing on offering a one-on-one experience, we empower users to practice at their own pace, allowing them to gain confidence and improve their speech skills without the pressure of doing it in front of others. Perfect Speech AI bridges communication gaps and creates an inclusive environment where everyone can improve and succeed.
What it does PerfectSpeech AI is an audio transcription tool that listens to and converts spoken language into written text. Whether you're recording a lecture, meeting, or any other conversation, this tool captures your speech and accurately transcribes it. It’s designed to save time and effort for anyone who needs quick and reliable transcriptions.
How w built it We built the project using a combination of Java for the backend, utilizing the Google Cloud Speech API to handle the transcription process. The frontend was designed with user-friendliness in mind, making it simple for anyone to upload an audio file, record directly, or retrieve their transcriptions easily. The backend processes the audio data, communicates with the Google Cloud API, and returns the transcription in real-time.
Challenges I ran into One of the main challenges I encountered was integrating the Google Cloud Speech API into the Java environment. There were also some initial issues with setting up the environment and making sure all dependencies were correctly configured. Since this was my first time working with speech-to-text functionality, I had to spend a lot of time troubleshooting and learning how to handle audio data effectively.
I’m incredibly proud of developing a fully functional speech-to-text system that works seamlessly. Despite facing initial technical hurdles, I was able to successfully integrate the backend with Google Cloud services and ensure that the audio transcription process runs reliably. Getting everything set up and pushing the project live marked a significant milestone for us. This is especially meaningful for Jaeden and me, as sophomores and first-time hackathon participants. Our hard work, persistence, and dedication to the project speak volumes about our passion and commitment to making a positive impact.
What I learned I learned a lot about working with APIs, particularly Google Cloud's Speech API, and how to manage dependencies and handle audio data in a coding environment. I also learned how to effectively build and structure a project that communicates between the frontend and backend, and how to integrate real-world services into a working product.
What's next for PerfectSpeech AI In the future, I plan to improve PerfectSpeech AI by adding support for multiple languages, refining the transcription accuracy, and making it more intuitive. I also hope to integrate more advanced features like real-time transcription for live conversations or meetings. As more people find it useful, I’d love to scale it and make it more robust for different types of users and use cases.
