Comunidad

Inspiration

What inspired us was our disappointment when Duolingo, the popular language learning app, removed community features that had helped many learners. This motivated us to create Comunidad, with a focus on fostering communication between Spanish speakers and non-Spanish speakers, so learners can grow and become a part of that community.

What it does

Comunidad is a language learning platform that combines structured lessons — similar to Duolingo — with integrated discussion boards, so learners can practice in a real community context. Users progress through lessons while being able to ask questions, share tips, and converse with native Spanish speakers at every step of their journey.

How we built it

We built LangLearning using the Gemini CLI, leveraging Google's Gemini AI to power language feedback, lesson generation, and conversational features. This allowed us to rapidly prototype and iterate on core features.

Challenges we ran into

One of our biggest challenges was selecting the right Gemini model for each use case. We had to carefully balance performance and speed — defaulting to a lighter, faster model for most interactions, while switching to a more powerful model when deeper language understanding or generation was required.

Accomplishments that we're proud of

We're proud of the learning path progression system. Seeing users move through a structured, adaptive journey with clear milestones reflects the core mission of LangLearning.

What we learned

This project taught us time management under pressure and how to make the most of cloud-based tools. We gained hands-on experience balancing speed of development with thoughtful decision-making when integrating AI services.

What's next for LangLearning

We plan to deepen the connection between discussion boards and learning units so conversations are contextually linked to active lessons, making community interactions more relevant and creating a tighter feedback loop between content and real-world practice.

Built With

• node.js
• prostgresql
• vite