Algorithmic risk scoring

The signal is
what it’s next to.

Adjacia sees new infrastructure the moment it’s certified, scores it by the company it keeps, and hands you the entire malicious cohort — before it’s ever used in an attack.

Today

Investigation

An analyst starts from one bad indicator and pivots, hop by hop, hoping to find the rest. It’s manual, it’s slow, and it happens after the attack.

With Adjacia

Discovery

Start from the whole stream of new infrastructure. Adjacia surfaces every domain adjacent to known-bad — the full cohort, scored and explained — the moment it appears.

How it works

A pipeline, not a lookup.

  1. 01

    Observe

    Every certificate issued in the public transparency logs is a new domain, the instant it’s born. Adjacia watches that stream as its forcing function.

  2. 02

    Correlate

    It links each domain to known infrastructure by the rare signals that mean something — dedicated IPs, networks, certificates, favicons — and discards the generic hosting everyone shares.

  3. 03

    Score

    A model reads the assembled neighborhood and returns a 0–100 risk score with reasons you can audit — not a black-box verdict.

Risk bands

One number, six tiers.

Confirmed-bad is a fact (100). Everything below it is inferred by adjacency and weighed by the model.

0–19Benign
20–39Low
40–59Suspicious
60–79High
80–99Critical
100Known bad

Born-on-the-wire

Certificate Transparency as a forcing function — you see domains at issuance, not after a victim does.

Guilt by association

One known-bad seed propagates across shared, specific infrastructure to the whole ring.

Rarity-weighted

Shared CDNs and budget nameservers are suppressed, so legitimate neighbors don’t drag innocents in.

Auditable

Every score ships with the reasons and the linking signals behind it.

See it score a domain. Or an IP. Or a subdomain.

Type any entity and watch its cohort assemble in seconds.

Open the live demo