Detection that learns your environment
Signature-based tools catch what someone has already seen. AI-MDR Essentials establishes what normal looks like in your environment, then surfaces the deviations — which is how novel activity gets caught before it has a name.
Two honest caveats up front: it needs a baseline period before it is worth anything, and every escalation still passes a human analyst before it reaches you. The model proposes; the analyst decides.
What you get
- Behavioural baseline — a learning period establishes normal patterns for users, devices and data flows.
- Anomaly detection — deviations are surfaced and scored, catching what signature tooling misses.
- 24/7 analyst oversight — every escalation reviewed by a person before it reaches you.
- Weekly reporting — anomalies raised, investigated and dismissed, with the reasoning shown so you can judge the quality of the triage.
- Monthly model review — what the model learned, what it is still noisy about, what we tuned.
- Guided containment — confirmed findings come with ordered, environment-specific steps.
What is not included
- Autonomous containment executed in your environment — that is AI-MDR Complete.
- Meaningful detection quality during the baseline period. The first two weeks are deliberately noisy; that is how the method works.
- Remediation labour, rebuilds and restores.
Your first 30 days
- Day 0–2 — kickoff; telemetry sources identified and scope confirmed.
- Day 3–7 — endpoint, identity, email and network sources connected.
- Day 8–21 — extended baseline. Longer than a signature deployment because the model needs the data.
- Day 22–26 — first tuning review; we show you what was flagged and what we suppressed.
- Day 27–30 — steady state and first full weekly report.
Billed monthly. No minimum term. Cancel any time before your next renewal date.