Yandori

What it does

Yandori is a real-time system that tracks how news spreads across 200,000 websites. It processes continuous information streams from RSS/Atom feeds, compressing them into fixed-size world state embeddings using a novel neural architecture. The system predicts future news states at multiple time horizons (1 minute, 1 hour, 1 day) and maintains a constant memory footprint of ~14 MB regardless of deployment duration.

Who it is for

Yandori is designed for researchers, journalists, media analysts, and developers who need to monitor and understand the propagation of news across the web in real time. It is particularly useful for those studying information diffusion, detecting emerging trends, or tracking the lifecycle of news stories.

Why it matters

Traditional news monitoring tools struggle with the sheer volume and velocity of online content. Yandori's approach compresses vast streams of articles into compact, predictive representations, enabling efficient analysis of news spread without storing all raw data. Its self-supervised importance weighting mechanism automatically identifies predictively relevant content without manual annotation, making it scalable and domain-agnostic.

Launch signal

Yandori was launched as a Show HN on Hacker News, indicating an early-stage product seeking feedback. The system is operational and tracks 200k websites, but detailed information about funding, team size, or commercial availability is not provided in the available sources.

Brand and naming

The name "Yandori" is short, distinctive, and easy to remember. It does not directly describe the product's function, which may help with brand recognition but requires marketing to convey its purpose. The brand positions itself as a research-driven tool, as evidenced by the accompanying arXiv preprint detailing the underlying technology.