Hi HN, I'm Aman, a 20-year-old builder from Nepal, and for the last few months my co-founder (Alok) and I have been working on a small but meaningful experiment: Can AI generate clean, usable mobile app UI screens instantly from a text prompt? Today, we’re releasing the beta of Elaric, a tool that does exactly that. What it does (current version) Right now, Elaric generates UI screens only — but across almost all major categories: Food delivery E-commerce Booking apps Finance Fitness Education Social Hospitality Ride-hailing Marketplaces You type: “Generate a home page for a food delivery app with categories, a search bar, and a featured section.” Elaric outputs a complete UI screen in seconds. You can generate unlimited screens and explore them visually. Why we built this In Nepal, building an app UI takes days or weeks. Figma is powerful, but for non-designers, it’s still overwhelming. We wanted something simple: “Describe the screen you want → get a usable mockup immediately.” We know this is not new at a global level, but it is new in our context — and we wanted to see how far two young builders with almost no resources can push this idea. How we built it We bootstrapped the entire thing with ₹51,125. No investors, no team, no external help. We trained and fine-tuned models on a lot of manually prepared UI structures and category patterns, and then built a rendering engine that constructs the final screen layout. It’s definitely not perfect — but it works surprisingly well for many categories. Why we’re posting on HN Because this community is where we’ve learned everything. We respect the honesty here more than any other platform. We want feedback on: What would make this actually useful for developers/designers? Should we prioritize code export or screen linking next? Should the editor come before flow generation? Any obvious limitations you see from the demo? Any categories we should add next? Live Beta: https://www.elaric.ai/ (If it crashes, we’ll scale quickly — our servers are tiny.) What’s next Based on early user requests, our next steps are: Add Figma export Add screen-to-screen linking Create a small built-in UI editor Generate React Native code Build full app flows Release an API for bulk UI generation Why this matters to us We grew up in a place where people genuinely believe “big tech can’t be built in Nepal.” Posting this here is a small step toward challenging that. If anyone here finds this useful — even for 10 seconds — it will mean everything to us. I’ll be here in the comments all day. Thank you for your time, and happy to answer anything. Aman Co-founder, Elaric https://www.elaric.ai/

Thanks everyone for checking this out! I’m here for the entire day and will respond to all feedback, questions, and criticism. Please feel free to be brutally honest — it really helps us improve.

We’re two founders from Nepal building a tool that auto-generates mobile app UI screens. It’s very early — only UI works for now. Would love to hear what HN thinks we should build next.

Happy to answer any questions about how it works. Right now it only generates mobile app UI screens (no full app logic yet), so we’re trying to understand what direction would be most useful for builders here. Open to any feedback.