-
-
Logo
💡 Inspiration As someone passionate about learning and language, I’ve always felt traditional vocabulary apps fell short. They often give you a dry definition and move on — no context, no depth, no memory hooks. When I started exploring LLMs, I realized there was an opportunity to build something smarter: a tool that teaches words and terms the way our brains actually learn — in layers, through stories, analogies, and real examples. That was the spark for SmartLearn AI.
⚙️ What it does SmartLearn AI is my attempt to reimagine vocabulary and terminology learning using AI. It breaks down complex words into simple, digestible layers — from basic definitions and mnemonics to idiomatic usage, analogies, and even academic or cross-disciplinary insights. You can also ask it custom questions like “How is ‘resilient’ used in psychology today?” and get contextual answers in real time. It’s designed to be flexible, intuitive, and open to everyone — no signup needed.
🛠️ How I built it I built the frontend using Next.js and TailwindCSS to keep it responsive and clean. The backend is written in Go, which gave me performance and structure. I integrated the Perplexity Sonar API and OpenAI to generate multi-level content and contextual Q&A. Everything is containerized with Docker and deployed on Google Cloud Run, so it scales easily and stays cost-effective.
🚧 Challenges I ran into The biggest challenge was designing prompts that could produce explanations at different complexity levels — what works for a 12-year-old doesn't work for a university student. I also had to make sure the UI stayed clear even with a lot of information on screen. Handling API limits and making everything fast, stable, and reliable was another layer of work behind the scenes.
🏆 Accomplishments I’m proud of I’m proud that I managed to ship an MVP that actually feels smart. It’s public, polished, and people can just start learning without jumping through hoops. The live Q&A works. The explanations scale in complexity. And most importantly, it shows how powerful and personal AI-driven education can be when it’s designed with learning in mind.
📚 What I learned I learned a ton about prompt engineering, especially for educational content. I saw how powerful Go can be for fast, reliable backend services. And I realized that even solo, with the right tools and focus, it’s possible to build something meaningful in a short time. User experience matters — and so does making things feel intuitive and friendly.
🚀 What’s next for SmartLearn AI Next, I want to add personalized learning — saved words, spaced repetition, and progress tracking. I’m also planning to expand into technical terminology across fields like medicine or law. A mobile-first version is coming soon, and I’m looking into adding offline capabilities. Eventually, I’d love to see SmartLearn AI integrated into classrooms, study apps, or even language learning tools.
Log in or sign up for Devpost to join the conversation.