This is the final product
trying to see if the Backend was working with the Frontend
First Sketch of the Idea
First time we managed to show the markers of the news in each city

Inspiration

We wanted to build a news experience that feels local, global, and trustworthy at the same time. With so much misinformation and overwhelming feeds, people often don’t know which headlines to believe. Our goal was to combine interactive visualization with AI-driven transparency so users can quickly see news from their city or across the globe and understand why it’s credible.

What it does

Global News App lets users:

Explore a 3D interactive globe to zoom into a city of interest. View a live local feed of stories with credibility scores. Read AI-generated summaries for fast context. Check a Transparency Dashboard that shows how agents fact-check and cross-reference stories This makes news consumption faster, more engaging, and more trustworthy.

How we built it

Frontend (MVP First):

Built an interactive globe with Three.js for navigation. Created a clean news feed with React + TailwindCSS, starting with mock data.

Backend & Database:

Developed APIs with Express.js. Stored news posts (title, summary, source, geotag, credibility score) in MongoDB.

AI Agents:

Summarizer Agent (LLM) → creates TL;DR versions of articles. Fact-Check Agent → validates headlines against trusted sources (e.g., Google News API).

Challenges we ran into

Balancing hackathon speed with feature complexity (building a globe UI + AI pipeline in <48 hrs). Designing a credibility scoring system that feels transparent but still lightweight. Debugging rendering issues with Three.js and keeping performance smooth on lower-end devices. Coordinating backend and frontend work under tight deadlines.

Accomplishments that we're proud of

Built a working globe-to-feed flow that feels polished and demo-ready. Integrated AI agents that provide real-time summarization and fact-checking. Delivered a national-scale project in hackathon time while keeping the UI engaging and simple. Showcased a transparency layer that makes the AI process understandable, not a “black box.”

What we learned

How to quickly prototype with Three.js and React under time pressure. The importance of starting with visual MVPs before backend complexity. How to design simple but effective multi-agent workflows. That transparency in AI is just as important as accuracy when it comes to trust.

What's next for Global News Network

User posting flow: allow communities to contribute their own stories. More AI agents: e.g., bias detection, language translation for global accessibility. Mobile version: with gesture-based navigation and push notifications for local updates. Partnerships with news outlets to provide verified feeds at scale.