ReLang

Inspiration

Our inspiration derived from firsthand experiences, observing the common challenges individuals face when trying to improve their English grammar. We recognized a significant gap in accessible, interactive learning tools that not only correct grammar but also engage users in meaningful conversations. This realization led us to develop an AI bot designed to converse with users, providing personalized feedback and encouragement.

What it does

Our project helps the user hone their grammar skills through conversation with an AI bot. Our bot aims to create a supportive learning environment, helping users correct their grammatical errors. Through continuous interaction and practice, users can gain confidence in their English proficiency while receiving immediate corrections and suggestions.

How we built it

We built our frontend with Figma and Swift to create an appealing and intuitive user experience. Our backend handles API requests and uses that information for a grammar checker model. Based on the results from the grammar checker, we feed that information to a second model to generate responses. These responses can be congratulatory messages for excellent grammar or words of encouragement. The data flows from Speech-to-Text, then to Text-to-Text, and finally to Text-to-Speech. This creates a unique conversational environment that enhances user engagement and learning.

Challenges we ran into

For the backend, we had difficulties implementing openVINO on non-exported models. This caused a lot of wasted time. Instead of being stubborn and trying to get the code to work, it was more efficient to find models already made in openVINO.

Accomplishments that we're proud of

We designed and implemented an interactive AI bot that engages users in dialogue, helping them practice and improve their English grammar skills. Throughout the project, we enhanced our technical skills in software development, design tools, AI implementation tools, and as well as our soft skills in communication and problem-solving.

What we learned

We learned to work with FIGMA, AI models, and Intel software, as well as soft skills like effective communication, teamwork, and problem-solving. These soft skills were essential in collaborating with our team to brainstorm ideas and develop solutions for our grammar AI bot. Effective communication allowed us to articulate our vision and ensure that everyone was on the same page, while teamwork fostered a supportive environment where we could leverage each other's strengths.