Hivewire is a news app that lets you define what you want to read about, rather than inferring it from your behavior. We process thousands of articles daily from hundreds of sources and rank them based on explicit preferences you set.

How it works:

• Instead of collaborative filtering or engagement-driven ranking, you assign weights across four levels (Focus, More, Less, Avoid) and the engine prioritizes the intersection of your high-weight topics while aggressively down-weighting what you don't care about.

• Articles are clustered by story so you get one entry per development, not 15 versions of the same headline.

• Every morning, it pulls your top clusters and uses an LLM to generate a narrative briefing that summarizes what matters to you, delivered to your email.

Currently web-only and English-language. We'd love feedback from the community on the relevance of feed results, the UI, and the quality of the clustering.