Inspiration

When I first entered the workforce, my biggest challenge wasn't my technical ability—it was communicating my ideas with confidence. I often struggled to build trust with more senior members of my team because I didn't feel confident in the way I spoke.

At first, I assumed my foreign accent was the problem. Over time, I realized it wasn't just pronunciation. My rhythm, stress, and intonation played a major role in how confident and natural I sounded.

That led me to look for ways to improve, and I discovered shadowing: a practice technique where you choose a speaker and imitate not only their words, but also their rhythm, pauses, stress, and intonation. Research has shown that shadowing can be an effective tool for improving spoken fluency and prosody.

I wanted a way to practice shadowing while getting concrete feedback on what I was doing differently from the speaker. That's what inspired RhythmFlow.

What it does

RhythmFlow lets you choose a speaker you want to sound more like by uploading an audio clip or providing a video.

The app breaks their speech into phrases, identifies where the main stresses and pauses occur, and lets you record yourself shadowing the speaker.

It then compares your recording with the reference and gives you actionable feedback on your rhythm, stress, pauses, and intonation so you know exactly what to work on.

How we built it

I built it using lovable to design the front-end and Claude code and Supabase to integrate the back-end and create the algorithm to analyze the audio.

Challenges we ran into

One challenge I ran into was making the algorithm that analyses the audio more reliable. I was having issues with how it breaks the clips into phrases. The current algorithm occasionally segments phrases incorrectly, which can affect the feedback it generates.

Accomplishments that we're proud of

I'm very proud of the user interface, it's exactly what I wish I had when I started working on my communication.

What we learned

I learned that pronunciation of words is just one part of the equation of how people perceive you. Rhythm, stress and intonation are equally as important

What's next for RhythmFlow

I still need to fine-tune the algorithm that analyses the clips. I'd like to publish the app on the App Store!

Built With

Share this project:

Updates

Submission history