Security that learns your normal

Catch the attacks nobody wrote a rule for.

Rules only catch what someone has already described. Squirks learns what normal looks like in your own data and puts the few events that don’t fit at the top of the list. You don’t have to guess what attackers will try next.

Real case: an attack on our own gateway 8 sec to the top of the list 0 rules involved 1 line to read Read it →

How it works

  1. Point Squirks at your data

    Logs, network records, equipment output: any data, as it is. Nobody has to explain the format.

  2. It learns your normal

    Squirks builds a baseline from your own data. No rules, no threat lists, no assumptions.

  3. You read the short list

    Every event is measured against that baseline. The few that don’t fit rise to the top, grouped so lookalikes are one line.

Pay for decisions, not gigabytes

Cost follows the work, not how much you keep, so you can watch more.

Nobody writes the rules

Squirks learns normal from your data. There is nothing to author or maintain.

The strangest events come first

Ranked by how far they sit from normal, not by when they arrived.

Who it’s for

Teams responsible for more systems than they can afford to watch closely, including the equipment and appliances that ordinary security tools can’t read. If you have data and not enough people to read it, Squirks decides what they should read first.