Trust Stack / Intelligence

Real-Time Threat
Scoring

Adaptive analysis identifies unusual communication activity and supports policy-driven response with traceable decision context.

Threat Signals We Analyze

Every message and session is evaluated against multiple threat vectors before a policy decision is made.

Bot Behavior Detection

Identifies automated traffic, credential stuffing, and scripted abuse patterns.

Phishing Detection

Analyzes message content, URLs, and sender patterns for phishing indicators.

Anomaly Detection

Flags unusual volume spikes, destination patterns, and timing anomalies.

Pattern Recognition

ML models learn normal behavior per account and flag deviations in real-time.

Impersonation Risk

Detects sender name spoofing, brand impersonation, and social engineering attempts.

Replay Attack Prevention

Identifies repeated or reused message payloads and suspicious retry patterns.

How Risk Scores Are Calculated

Multiple signals are weighted and combined into a real-time risk score per message.

Sender reputation history

Past delivery rates, complaint ratios, and carrier block history.

High
Content risk analysis

NLP-powered scan for spam keywords, phishing links, and suspicious patterns.

High
Device & session signals

Device fingerprint, IP reputation, and session behavior anomalies.

Medium
Destination risk

Country-level fraud rates, DND registry status, and carrier block lists.

Medium
Volume patterns

Sudden spikes, unusual hours, and deviation from established sending patterns.

Medium
Recipient engagement

Historical open rates, opt-out velocity, and complaint trends.

Low

Automatic Response Actions

Based on the risk score, the system takes proportionate action in real-time.

Low
Allow

Message routes normally through carrier networks.

Medium
Review

Message flagged for monitoring. Sent with enhanced logging.

High
Throttle

Rate-limited. Sender notified. Manual review recommended.

Critical
Block

Message blocked. Account suspended pending investigation.

Decision Flow

From message submission to policy action — fully automated, fully auditable.

01

Observe Context

Collect sender identity, device signals, content analysis, and destination risk data.

02

Assess Risk

ML models evaluate all signals against learned patterns and configured thresholds.

03

Guide Response

Policy engine applies proportionate action and records the decision rationale.

Intelligent Security, Transparent Decisions

Every risk score is explainable. Every action is auditable. Every decision retains its context for governance review.