SpeakSquid

Elevate your language learning with AI-powered practice, offering custom prompts and preset scenarios in 57 languages, while tracking your progress through immersive video-based speaking exercises.

Comment

Inspiration

Around 1.2 billion people are learning a new language right now, but unfortunately the average cost of language lessons can range from $20-$50 per hour and mastering a language outside its native environment can be challenging. Our project SpeakSquid was made to overcome these obstacles, providing free and personalized speaking practice for language learners, focusing on three pivotal factors of language learning: Speaking, Correction, and Motivation.

What it does

SpeakSquid tackles the most important aspect of learning a language—speaking. With SpeakSquid, users are able to practice speaking in the language they are learning to an AI with personalized scenarios using custom prompts while recieving grammatical corrections and sentence analysis for future improvement.

How we built it

For this project, our tech stack includes HTML, CSS, JavaScript. We used Node.js, Express and Passport for the login/register page and OpenAI’s Whisper API for the speech to text transcription, the GPT API, Speech Synthesis API for the text to speech, and the Libre Translate API for the translations in the transcript.

Accomplishments that we're proud of

We are proud of all the features we challenged ourselves to use for this hackathon. We had a lot of big visions for this project such as having 10 pre-set scenarios for people to practice speaking with before making their own prompts, and adding a transcript that corrects grammar, and we are super happy we were able to implement everything we wanted to!!

What we learned

We learned how to implement many new APIs such as Speech Synthesis API and Libre Translate API as well as embed a camera with functioning camera and microphone components. We also learned how to more efficiently work together through sequencing the parts of the projects we were working on at the same time to fit together and be able to swap parts efficiently.

What's next for SpeakSquid

SpeakSquid has the potential to be big for language learners worldwide. While our application is working and ready for use, we have tons of ideas to enhance our product. For example, we hope to eventually add a pronunciation assessment API. Currently, the ones we found all require payment to use so we didn’t implement it, although with enough users we could potentially look into it. We hope SpeakSquid will be able to help empower users to gain confidence in their speaking when learning new languages. Together, let's unlock the power of your voice!!

Built With

Share this project:

Updates