Features

Sentinel combines several independent detection mechanisms into one risk decision. You don't need to enable everything — most deployments start with the defaults and add features as their use case requires.

Core detection

• Adaptive Captcha — scales what a visitor sees to their risk score, from a frictionless proof-of-work check up to an interactive Code Challenge, instead of a fixed difficulty for everyone.
• Human Interaction Signature (HIS) — interaction analysis that estimates the probability a request came from a human.
• Autopilot — automatically tunes challenge difficulty and policy based on observed traffic, instead of a static config you set once.
• Threat Intelligence — real-time IP reputation and blocking from open-source and live-traffic sources.

Content moderation

• Classifier — NLP-based spam and content scoring for submitted text.
• Phishing Detection — validates URLs in submitted content against known phishing sources.
• Similarity & Training Data — pattern-matching against example datasets you provide, for custom moderation rules.
• Email Spam Filter — spam scoring for inbound email.

Abuse prevention

• Rate Limiters — flexible, multi-layer throttling by IP, key, or custom identifiers.
• Detection Signals — supporting signals (IP resolution, header entropy, language detection, link/redirect analysis) that feed into the overall risk score.

Where to start

If you're not sure what to enable first, see Use Cases for recommended feature combinations by scenario, or just leave Autopilot on and let Sentinel tune itself against your real traffic.