KQL Thinking for Analysts
Move from an alert entity to useful pivots instead of searching randomly.
Practical investigations, detection techniques, threat hunting and security research across the tools analysts use every day.
Move from an alert entity to useful pivots instead of searching randomly.
Build a search path from source, user, host and time to correlated activity.
Work from offense context to contributing events, assets, users and evidence.
Process trees, alerts, users, network connections and device timeline pivots.
Trace process ancestry, command lines, detections and related endpoint activity.
Use endpoint inventory and activity context to validate scope and exposure.
Use indicators as leads and behavior as context.
Start with a behavior hypothesis and identify the telemetry that can test it.
Correlate command lines, parent processes, users and network activity.
Look for authentication, remote services and unusual host-to-host patterns.
Turn observable behavior into portable detection logic.
Reduce false positives without removing meaningful coverage.
Connect detections to adversary techniques and coverage gaps.
User, device, location, authentication method, risk state and follow-on activity.
Source IP, targeted users, failure patterns and successful authentication.
Authentication, MFA, sessions, privilege and post-compromise activity.
Authentication, process, account and object-access telemetry.
Parent-child relationships, command lines and execution context.
Scheduled tasks, services, startup locations and other persistence points.
Correlate events across identity, endpoint and network telemetry.
Metadata, strings, imports, sections and indicators without execution.
Headers, sections, imports and suspicious characteristics of Windows binaries.
Processes, files, registry, DNS, network connections and runtime behavior.