About the Project: AI Good-News Radar

Inspiration

The spark for this project came from a deeply personal place. I noticed that my wife (while being 6 months pregnant) start her day scrolling through news feeds filled almost entirely with negativity. As someone who wants nothing but the best energy around my wife, especially going through our first pregnancy together, it hit me how something as silly as a negative story can shape your outlook for the day. I wanted to create something that flips that experience: an app that surfaces only positive news, backed by real data, so people can begin their day feeling hopeful instead of drained. It started as a simple wish to bring a little light into our mornings, but I realized it could help so many others too.

What I Learned

Working on this project opened my eyes to the real power of AI—especially sentiment analysis—not just as a technical tool but as a force for good. I saw how AI can sift through overwhelming amounts of data and find the bright spots that might otherwise get lost. It’s incredible how technology that’s often seen as cold or impersonal can actually help us connect with the positive side of our world. This made me believe that AI, when used thoughtfully, can amplify good deeds, spread hope, and remind us that there’s more light than darkness if we just look for it. A few more learnings:

• The challenges of pulling together thousands of news sources and turning that chaos into clear, uplifting themes that people actually want to read.
  
• Designing a clean, warm, and approachable user experience that doesn’t overwhelm but invites daily engagement.
  

How I Built It

• I started by exploring different ways to gather positive news, initially experimenting with APIs like NewsAPI and scouring Reddit feeds for uplifting stories. It quickly became clear that finding reliable, consistently positive content required a more curated approach, so I shifted to pulling in RSS feeds from trusted, positivity-focused sites.
  
• From there, I integrated AI-powered sentiment analysis to sift through thousands of headlines daily, helping the app distinguish genuinely uplifting stories from the noise. I then built a clustering system to group related stories into themes—so instead of random headlines, users get meaningful collections like “Clean Energy Wins” or “Acts of Kindness.”
  
• To make the experience engaging and intuitive, I designed a live radar-style map that visually highlights global positivity hotspots in real time. The mood tracking feature was intentionally kept simple—a thumbs-up or down—so users can easily reflect on their feelings and see how they align with the world’s mood.
  
• Finally, I developed a predictive model to forecast which positivity themes might trend next, giving users a sneak peek into tomorrow’s good news. Throughout, I focused on a warm, minimalist mobile-first design inspired by apps that feel calm and inviting, making sure the tech didn’t overwhelm the message but amplified it.

Challenges

• Handling inconsistent and messy RSS data required extensive cleaning to ensure reliable sentiment analysis.
  
• Balancing AI complexity with app performance to keep the experience fast and smooth.
  
• Designing an uplifting yet simple interface without crossing into cheesy or overwhelming territory.
  
• Protecting user privacy while delivering meaningful global positivity insights through anonymous, lightweight data handling.

This project reminded me how technology can—and should—be used to make people feel better, not worse. It’s about creating small moments of hope and connection in a noisy world.

Built With

• bolt
• react
• typescript