Conversify

Home Page
Conversation bot page (no transcript)
Conversation bot page (with transcript)

Inspiration

As a team with diverse backgrounds and experiences, we were inspired by the power of languages to connect people across cultures. We saw an opportunity to harness this potential through technology, drawing inspiration from tools like ChatGPT to create a solution that makes language learning engaging and accessible for everyone.

What it does

Our web app, Conversify, is a language learning tool designed to simulate real-world conversations. It allows users to choose a language and proficiency level, then engage in a dynamic dialogue with a machine learning model. The app provides instant feedback through audio playback and a transcript, helping users practice and improve their language skills in a natural and interactive way.

How we built it

On the frontend, we used React, incorporating asynchronous state management to handle dynamic audio data flows. Browser web APIs were employed to process the audio input we receive from the user in real-time, which was then communicated with our Flask backend. Our Flask-based REST API orchestrates data exchange between multiple endpoints, allowing for sophisticated tasks like AI-driven text generation via Cohere's Large Language Model (LLM). We also integrated React Router for efficient route management and a SpeechSynthesis API to deliver real-time, natural-sounding text-to-speech conversion, enhancing accessibility and user engagement. This allowed our app to have conversations with users in any given language.

Challenges we ran into

Lots of merge conflicts
Debugging asynchronous calls
Audio conversions
Styling issues

Accomplishments that we're proud of

Our team's proud accomplishments include integrating advanced audio processing, designing an engaging and intuitive UI, and ensuring smooth data transfer between the frontend and backend. This project has also been an invaluable learning experience, equipping us with lessons that will benefit us in future endeavors.

What we learned

Making an API
Routing
Dynamic behaviour
Audio analysis

What's next for Conversify

For the future of Conversify, we plan to expand the range of available languages, offering users more options for practice. Additionally, we aim to develop a grading system to provide feedback, comparing users' performance with others at the same language and proficiency level. This will help track progress and encourage continued improvement.