Retina

Retina is the lab’s news perception. It is built embedding-first: every article is turned into a vector before anything else happens, so that similarity, clustering and search all work on meaning rather than keywords. A language model then extracts the entities each article mentions (people, places, organisations, events, assets) at a measured precision of 95.7 %, and links them to a curated catalogue. Storylines are clusters of articles that stay together over days; bursts are names that suddenly outrun their own baseline; tone is a sentiment score per article, aggregated per entity.

Sources are ordinary: RSS from newspapers, broadcasters, agencies and specialist outlets, a set of public Telegram channels, and a few sources that are only reachable over Tor, routed through the lab’s own gateway. Nothing is scraped that a publisher hides behind a login. Retina keeps derived data; the articles stay with their publishers.

Retina replaced a much larger predecessor on 26 March 2026: fifty-seven containers and more than two hundred tools, retired in one move for a single service with one job. The 96 % did not come for free either. Until July 2026 articles were linked to entities by vector similarity, and when that was measured it reached 2.4 % precision on a test name. The language-model extraction had been running alongside all the time, its results thrown away; since 18 July it is the only source of the links. Consent, login and bot pages are recognised before any request is made and dropped: about 5 % of the inflow, an extrapolation from the first afternoon and a check of 78 cases; the seven-day measurement runs until 8 September.

The numbers below are read from the running system on the day of the snapshot. Retina feeds the actor graph, the situation picture and the network on the front page.