How Vantage works
Vantage collects technology stories, scores how widely each one is being discussed, and writes it up. It runs on a schedule with nobody in the loop.
It reads a spread of sources
Wire services, developer communities, builder forums and regional publications. The point of the spread is that the input includes what engineers and founders are discussing, alongside what reporters have filed.
Stories are scored by how widely they are covered
Each story is weighted by how many independent sources are carrying it and how much engagement those sources are getting. Something running across several communities scores higher than a single press release.
Each story is written up in full
An article covers what happened, who it affects and what to watch next, rather than a reworded headline. The write-up is generated from the sources, so it is a starting point: follow the links before relying on it.
Six regions, framed locally
Coverage spans six regions, and each story is framed for where it comes from, whether that is fintech in Nairobi, semiconductors in Seoul or regulation in Brussels.
What you can do
Ask questions on a story
Each article has a chat attached to it, so you can ask a follow-up or push back on what the analysis claims.
Search, or generate
If a story is not in the archive yet, searching for it writes the analysis on the spot.
Daily briefing
The day's stories in score order, meant to be skimmed rather than read through.
Save for later
Bookmark articles to come back to. No account needed.
Social pulse via Virlo
What is trending on TikTok, YouTube and Instagram, pulled from Virlo's API.
Why this exists
Wire services move quickly and stay shallow, opinion pieces are written for engagement, and most of both is aimed at one market. I wanted something that read several kinds of source at once and used the overlap between them as the weighting.
Where traditional media, developer communities and builder forums land on the same story, it is probably worth your attention. Where they disagree, the disagreement is usually the more interesting thing.
There are no editors and no sponsors here. It is also fully automated, which is the honest limitation: the analysis is only as good as the sources it happened to read, and it can be wrong.