Skip to content
Banner-Threat detection

Detection that runs where your logs live

Rules, baselines and anomaly scoring all run inside your own Guardsix deployment, with no cloud service or external model call in the path. Self-hosted and air-gappable, your detection logic and behavioural data stay under your jurisdiction.

Threat detection shouldn't wait for a team you don't have

Most detection programmes assume a detection engineering team to write rules, tune them and track coverage. Lean teams inherit the same expectations without the headcount, and the gaps stay hidden until an attacker finds them.
Protection starts weeks after go-live
Someone has to choose what matters, write it, test it and tune out the noise before real coverage begins.
Updates undo the tuning you did
Packaged detection updates can overwrite the thresholds and exclusions your team spent weeks getting right.
Coverage gaps stay out of sight
Without a clear map of how each rule detects, a simple question about coverage takes hours to answer.
Signatures miss what looks legitimate
Alert rules catch what you already know to look for. Unknown threat activity never trips a signature.

One threat detection system, not a stack of tools to keep in sync

Guardsix SIEM brings detection-as-code, guided rule discovery, a curated starting pack and explainable anomaly detection together as one system.

How the detection system works

Two lanes feed one outcome. The top lane covers known attacks with versioned, mapped and curated rules. The bottom lane learns what normal looks like in your estate and flags what departs from it.

Detect what you know

  1. 622 rules

    Detection-as-code

    • Every detection a versioned YAML file
    • Stable rule IDs keep your tuning intact
  2. Discovery

    Use Case Navigator

    • Exact query and requirements per rule
    • MITRE ATT&CK heatmap shows the gaps
  3. 63 rules

    Day-One Detection Pack

    • Curated from the full library
    • Live from day one, before you write a rule

And surface what you don't

Add-on for SIEM

Anomaly Detection

  • Flags behaviour that departs from a learned baseline
  • Every finding carries its reason and score
  • Anomalies can feed the same alert rules

One alert and incident stream

  • Richer leads, with signature and behavioural evidence together
  • Faster mean time to detect and respond, with no separate queue
  • An audit trail recorded as part of the work
Gradient dark mesh

What changes for your team

Each part removes work your team would otherwise do by hand.

TOUCH
Detection-as-code Your tuning survives every update
Detection-as-codeYour tuning survives every update

Versioned rules with stable IDs, so an update never overwrites your work.

TOUCH
Use Case Navigator See your coverage gaps at a glance
Use Case NavigatorSee your coverage gaps at a glance

Every rule's query and requirements, mapped to a MITRE ATT&CK heatmap.

TOUCH
Day-One Detection Pack Protected from the day you go live
Day-One Detection PackProtected from the day you go live

63 curated rules for the attack stages that matter most, running before you write one.

TOUCH
Anomaly Detection Catch what no rule describes
Anomaly DetectionCatch what no rule describes

Learns normal behaviour and explains every deviation it flags. The analyst keeps the decision.

TOUCH
One alert stream One queue with the full picture
One alert streamOne queue with the full picture

Signature and behavioural evidence arrive together as one richer lead.

TOUCH
Predictable cost A hard week never moves your bill
Predictable costA hard week never moves your bill

Priced on the infrastructure you run, with AI for SIEM inside the node-based licence.

Gradient dark yellow mesh

See threat detection running on your own terms

Get rules, baselines and anomaly scoring all run inside your own Guardsix deployment