Equalizer

• 

  Landing page

• 

  Recording page

• 

  Processing page

• 

  Results page

Inspiration

As a kid who grew up in an immigrant family, my parents' first languages are not English, and they struggled with pronunciation throughout my childhood. The area I saw this "broken" pronunciation affect the most was the world of work. Despite their extensive experience in their respective fields, it is undeniable that today's work world values communication over all else, and clear English speaking is key so securing a position.

After doing some research, I found out that those with standard, English, accents were hired significantly more as those with non standard accents. Many English as a second language learners look towards English language coaches to correct their pronunciation, but for many immigrant families, English language coaches are far too expensive and inflexible for their schedules. For example, my parents were unable into fit English coaching sessions into their busy and erratic job-searching schedules.

This inspired me to try to improve the accessibility of English pronunciation correction. I wanted to get rid of fixed schedules, high prices, and slow results so immigrant families like mine can thrive in North America and people searching for their jobs can utilize their full potential.

What it does

Equalizer is a web application made to provide detailed feedback for your English pronunciation. Record a clip of any length, of any content-during an interview, in a presentation, or just have it open throughout the day-and receive professional feedback to quickly improve your pronunciation. Find out exactly what you need to work on and receive actionable next steps for your speech.

In-Depth Analysis of Speech: Provides a score from 0-100 and detailed steps for improvement for Phoneme Accuracy, Fluency, Intonation, and Connections.

Quick Feedback: Equalizer can provide results within 10 seconds, ensuring instant progress.

Flexible and Accessible: Available for use at any time, in any place, making it much more flexible than sessions with language coaches.

How we built it

Frontend Framework: React. This is a single-page application (SPA) built in React for quick and responsive UI.

Backend Framework: FastAPI, Python. FastAPI efficiently communicates with the frontend.

AI/LLM Services: Google Gemini 2.5 Flash. This model is quick and efficient at processing audio.

Challenges we ran into

I had to learn how to record audio on the frontend, and passing that audio to the backend. Additionally, I did not realize at first that I had to convert the audio from the frontend into a format that the LLM could read (I chose .wav).

Accomplishments that we're proud of

Building a full, working website with a landing page and solving one of the issues I noticed as a child

Creating a nice GUI and colour palette

Implementing an advanced system of user input (audio)

Creating a complex pitch video in a the short time frame of one afternoon

What we learned

Web audio recording with React

Passing audio data to backend with FastAPI

What's next for Equalizer

Implementing a sign-in system to track progress (analysis results) over time, to allow users to see their improvement.

Built With

• fastapi
• python
• react