SnapNews

Making information faster and more accessible for the world

Comment

Inspiration

Everyone wants to stay up to date with what's happening in the world. First, we relied on newspapers. Then, people tuned in to radio stations for the news. It moved on to watching news on television, and then finally consuming news articles and videos on the internet. From paper, to radio, to TV, to internet - with each step, information become faster, more reliable, and more accessible. SnapNews aims to be the next step in that evolution, powered by the latest AI technologies.

What it does

SnapNews uses AI to deliver bite-sized summaries of what's happening around you, using information from hundreds of credible sources across the world, so you don't have to. What would have taken you minutes, or hours - in seconds.

How we built it

NextJS for the frontend and server-side rendering & caching. We also used flask for the backend, combined with a data pipeline using newsapi.org's news api, and OpenAI's GPT 3.5 to generate the summarized articles.

Challenges we ran into

Since most of the stack was new to all of us, we struggled initially with getting our app to look how we wanted it to. We ended up with a less feature-rich frontend, but we still have potential to improve it. Also, we planned to use dynamic loading, summarizing, and server-side caching of News that can be triggered from the user's side, for live refreshing. While this didn't work as intended, we still managed to create a pipeline that could be triggered at periodic intervals from the backend, to refresh news. And we did manage to get a working product in the end.

Accomplishments that we're proud of

We're happy we learnt to use an almost entirely new tech stack despite the many challenges that came with it, and that we created a fully automated summarizing workflow that would have taken dozens of man hours if done manually. Our project sets us up for a faster, more accessible future of news.

What we learned

We gained proficiency in diverse technologies, including Tailwind and ShadCN for frontend development, enabling us to create an interactive user experience. Additionally, we acquired skills in ethical web scraping, content summarization with ChatGPT, and image generation using DALL-E. Our collaboration improved through effective use of Github, highlighting the significance of branching.

What's next for SnapNews

We plan to develop a tinder-styled swipe card UI to users so we can learn what kind of news topics they're interested in, and make sure to show them the latest news in those topics first. We also plan to make a credibility-meter that can be shown next to news, using degree of information in common with other news sources covering the same information.

Built With

Share this project:

Updates