Multi-Model AI

Exaforce's three-model engine. The Semantic Data Model resolves entities and relationships, the Behavioral Model learns normal activity, and the Knowledge Model reasons across both.

Defined in the AI SOC glossary. Covered in depth on AI SOC architecture. More on exaforce.com.

Other terms

Agentic SOC — AI agents carry work to completion; analysts supervise.
AI SOC — The broad term, covering assistant through to full agent.
Alert triage — Deciding whether an alert is real. Where analyst hours go.
Exabots — Exaforce's five task-specific AI agents.
MDR — Managed detection and response, run by an external team.
MTTI — Mean time to investigate: how long to reach a verdict.
MTTR — Mean time to respond: detection through to containment.
Semantic context — Structured meaning attached to events at ingest.
SIEM — The traditional log aggregation layer. Exaforce is not one.
SOAR playbooks — Deterministic automation scripts, and why they break.
Threat hunting — Searching for adversary activity that no alert fired on.
UEBA — Baselining normal behavior to spot deviation from it.

All terms