Inspiration
During our brainstorming session for the hackathon, we explored various industries in search of challenges to tackle. One idea that caught our attention was the concept of an AI-powered bedtime story reader. However, as we delved deeper into this idea, our discussions led us to envision an AI language companion. This companion would engage in conversations with users, aiding them in learning a new language. Recognizing the scarcity of competitors in this niche, we decided to develop this concept further. We discussed the business model, pricing strategies, feature sets, and long-term plans, shaping not just a product but also envisioning the formation of "Lingua" – a "lang-tech" company.
What it does
Lingua Learner is not an AI tool but rather a companion, a friend from another country who can help you learn a new language. It's as simple as taking out your phone and talking to it. It simplifies the language learning process by enabling users to converse naturally, without the need for structured learning sessions or quizzes. We have designed Lingua Learner in such a manner that you can naturally converse with it, and talk to it about your hobbies, interests, and happenings in life. Through interactive dialogue, Lingua Learner gradually introduces users to their target language, providing feedback on fluency, pace, tone, and sentiment analysis. By focusing not only on what users say but also on how they say it, Lingua Learner facilitates learning language in everyday contexts.
How we built it
We decided to have a ReactJS frontend where our user will interact with Lingua Learner while our data processing, interacting with whisper API and openAI API along with the audio analysis will take place on the Flask backend. Our data is stored on Firebase. We also used an audio sentiment analysis to help the user.
Challenges we ran into
We ran into several challenges during the ideation and development of Lingua Learner. We ran into a roadblock in the beginning when we realized that we needed more differentiating factors from our competitors. A majority of our team comprises bilingual individuals and we realized that feedback on tone, pace, and diction is extremely important to be able to converse in a language, which is when we added audio analysis as a key component of our product.
Accomplishments that we're proud of
We take pride in successfully building an MVP within a tight timeframe of just 24 hours. Our achievement lies not only in facilitating language learning but also in implementing comprehensive audio analysis, a task that required extensive research and debugging. Lingua Learner stands out for its user-friendliness, catering to a wide demographic, including older users.
What we learned
Throughout this project, we gained valuable insights into audio file analysis and acquired a deeper understanding of various languages through testing and development.
What's next for Lingua Learner
In the future, we aim to enhance Lingua Learner by integrating additional features such as accent analysis and guidance, identifying common user mistakes, and further refining sentiment analysis across multiple parameters.
Log in or sign up for Devpost to join the conversation.