PulseAugur is the signal layer for AI news.
Every AI story gets covered by a dozen sources within hours. We cluster the coverage, score it on signal, and put the stories that actually matter at the top.
What PulseAugur is
PulseAugur is a daily intelligence layer for the AI industry. We ingest from 200+ sources — mainstream press, lab blogs, arXiv preprints, Hacker News, developer communities on Bluesky, Mastodon, Reddit, and X — and we cluster every story into a single page that consolidates the multi-source coverage of that one event. Each cluster carries a machine-written summary, a forward-looking AI impact blurb, a 0–100 newsworthiness score, the full source list with publishers and timestamps, and machine-translated titles for the non-English coverage.
The category PulseAugur fits into: AI-news aggregation. The shape it takes: cluster-per-story rather than link-per-source. What that means for readers: when OpenAI ships a model, you see one cluster page with the OpenAI announcement, the New York Times piece, the Stratechery follow-up, the arXiv paper that dropped alongside, the Hacker News thread, the Reddit reaction, and the Bluesky take from the lab people who actually built the thing — ranked by signal, with duplicates collapsed. Different shape from The Batch (weekly, single-author), Techmeme (general tech, headline-stack), or Smol AI / AI News (daily, Discord-summary).
How clustering works
Every fifteen minutes a pipeline polls our source list, dedups near-duplicate stories, scores each cluster on six dimensions (authority, cluster strength, originality, headline strength, buzz, velocity), gates research / commentary buckets behind hedge detectors and quality thresholds, and surfaces the top ~100 clusters per day to the Brief. The full method lives in our editorial standards and we explain the algorithm itself in how AI news clustering works. Everything you read on this site is either a cluster of human-written sources or an LLM-written summary of those sources — never editorial commentary generated from nothing.
Editorial transparency
Cluster summaries and AI impact blurbs are written by
a current frontier large-language model — Gemini 2.5 Pro,
Claude Opus 4.7, and GPT-5 are the three we rotate through
depending on the cluster's bucket and language. Every summary
links to every source via schema.org
NewsArticle.isBasedOn — so the citation graph is
machine-readable as well as human-visible. Marked with an
AI chip at the end of the
summary text. The
editorial standards page
covers what we do, what we don't, and how corrections work.
Founded by
PulseAugur was founded in 2026 by Chris Valentine — twenty-plus years as a CTO and software architect, with prior platforms that powered 50,000 sites, served 1B users, and drove 100%+ revenue lifts in e-commerce. Chris built PulseAugur because AI news is everywhere and signal is buried. The founder page has the full backstory.
Contact
Editorial corrections, press inquiries, source tips, partnership questions — one address: [email protected]. Social channels live on the founder page.