CherryPick/ 2026AI · CONTENT · LIVE
From summarizing news to making sense of it.
There's no shortage of news. Understanding why a story matters is the harder part.
CherryPick selects ten key stories each day and explains the background and context behind them.

01The Problem
There's no shortage of news, but few places briefly explain why today's stories matter.
02Rethinking how information should be presented.
CherryPick initially started as a service that delivered concise summaries of news articles.
During development, I began reconsidering how news content could be used responsibly, particularly in relation to copyright and content usage. This led me to rethink the product's direction.
Instead of focusing on summarizing individual articles, I redesigned CherryPick around selecting important stories and explaining the context behind them.
The shift also helped clarify the value I wanted CherryPick to provide.
03How It's Built
I built an LLM pipeline on Next.js and Supabase that extracts a fact sheet from each article and writes issue cards from those facts alone.
Article text isn't stored; original articles are linked through “Read more.”



- My Role
- Product Planning / UI & UX / Development / Deployment
- Tech Stack
- Next.js / TypeScript / Supabase / LLM API / Vercel
- Platform
- Web
- Status
- Live
- Year
- 2026
- Links
- Website ↗