RikRok

RikRok was inspired by how overwhelming it can feel to keep up with new AI research. I’m constantly reading papers, but discovery is still slow and inefficient, so I wanted a faster and more intuitive way to browse what’s new. I built a mobile app that works like Tinder for research papers: swipe through the latest arXiv entries, and each abstract is summarized into three simple lines using NVIDIA’s Nemotron model, so you can instantly understand the core idea. I learned a lot about building gesture-based interfaces in React Native, parsing arXiv’s XML data, designing effective summarization prompts, and managing persistent state for saved papers. A key challenge was making the swiping interactions feel smooth and natural while also keeping the app responsive as summaries were being generated. Another challenge was getting summaries to be short but still informative enough to decide whether a paper is worth reading in full. Overall, the project helped me explore how UI, interaction design, and AI summarization can make research discovery more accessible and enjoyable.

Built With

• react