Email Intelligence Overview

Email-based risk intelligence for detecting fraudulent registrations

Email Intelligence is the Titan product for email-based fraud detection. It analyses the email address provided during a registration session — evaluating handle patterns, domain reputation, MX infrastructure, and velocity signals — to produce a composite risk score and a set of triggered rules, so you can identify suspicious sign-ups and make better risk decisions before accounts are created.

What it does


Scores email risk — Every response includes a composite risk score (0–100) you can feed directly into your risk policies.

Evaluates domain reputation — Assesses the email domain's age, MX configuration, disposability, and association with known threat infrastructure.

Detects handle anomalies — Flags gibberish handles, enumerated patterns (e.g. user12345), LUDO patterns, and sub-address abuse.

Surfaces triggered rules — Returns the specific rules that fired (e.g. disposable_domain, high_velocity_email) so you can understand why a session is risky.

Tracks email velocity — Identifies email addresses or domains appearing at unusual rates across sessions.

Why it matters


Email Intelligence is designed to detect risk associated with email addresses, not just bots. Emails are easy to manipulate and free to create, making human-driven registration fraud a significant threat. Common uses include blocking fake account creation at scale, identifying disposable or programmatically generated email addresses, and reducing downstream abuse from fraudulent accounts.

How it fits with Titan

Email Intelligence follows the shared Titan model — a client integration collects session signals, and your backend receives the email risk data via the Verify response. See How Titan Works for the common flow.

Next steps

Set up Email Intelligence — Follow the shared Titan setup guide (client + server), then complete the Email Intelligence–specific enablement steps.

Email Intelligence Rules Documentation — Review the full ruleset and what each rule detects.