Every Identity Stack Has a Blind Spot
Your stack confirms the data is correct. It confirms the document is real. It confirms the device is recognized. It does not confirm the person filling out the form is the person whose data is being entered.
Stradum does.
The Problem
The Gap in Every Existing Stack
Identity fraud costs businesses $43 billion annually. Every organization dealing with identity-sensitive forms already has a stack. That stack has a blind spot it cannot close on its own.
Knowledge-Based Verification
Fails when the person knows the answers. Family members, insiders, and social engineering victims all pass KBA. Knowing the data is not the same as being the person.
Population-Level ML Models
Detect anomalies relative to a population, but cannot distinguish between two people who both behave "normally." A spouse looks completely normal to population models.
Device Fingerprinting
Identifies the machine, not the human. A shared household device, corporate network, or VPN renders these signals meaningless. Two people, one device — invisible to fingerprinting.
Stradum closes this gap by comparing behavior against the individual's own baseline — not a population, not a knowledge base, not a device. Every person has a unique behavioral signature when entering data. When someone else fills out a form with their information, they cannot replicate their motor patterns, even if they know every answer.
Integration
Three Steps. One Script Tag.
No SDK build step, no UI changes, no user-facing challenge. Drop one tag onto any page with a form.
Add one script tag
Three lines of config. Works with React, Vue, Angular, Next.js, or vanilla HTML. Under 25KB gzipped.
User fills out the form
Behavioral data is captured passively and invisibly. No CAPTCHA, no challenge, no friction. The user never knows Stradum is there.
Get behavioral match signals
Receive a match score, match status, confidence level, behavioral indicators, and plain-language reason codes via API and webhooks.
<script src="https://api.stradum.com/v1/stradum.min.js"></script>
<script>
const session = Stradum.init({
apiUrl: 'https://api.stradum.com',
apiKey: 'pk_live_your_key',
formId: '3a0c433f-951c-4e66-8644-e6ed1f9z4918'
});
</script>That's it. Behavioral capture begins automatically on page load.
Identity Lifecycle
Stronger With Every Interaction
Your system identifies the person. Stradum confirms it's still them — and that confirmation gets stronger every time they interact.
Baseline Captured
Behavioral signature recorded. Bot and impostor detection active immediately. Score and status delivered from the first interaction.
Profile Strengthens
Each session is compared against the growing baseline. Natural variation absorbed. Confidence increases.
Strong Behavioral Identity
High-confidence decisions. Familiar deviation fully active. The signature carries across every form this identity touches.
You already know who the person is. Stradum builds a behavioral profile behind that identity and answers the question your stack can't:
Is this the same human?
Output
What You Receive for Every Session
Signals designed to plug directly into your existing decisioning workflow. No changes to your current stack required.
Match Score
0-100 behavioral match score measuring how well the session aligns with the identity's established pattern.
Match Status
Three clear outcomes: match, review, or mismatch. No ambiguity.
Confidence Level
Tells you how much to trust the result, driven by baseline depth. Grows stronger with every session.
Behavioral Indicators
Five signal categories — Typing, Interaction, Navigation, Data Entry, Consistency — each rated expected, atypical, or anomalous.
Reason Codes
Human-readable explanations your analysts can act on. No internal methodology exposed.
Familiar Deviation
A unique signal: data entered with familiarity, but behavioral patterns inconsistent with the identity. Only Stradum detects this.
Detection
Three Layers That Work Together
From bots to insiders — each layer catches what the one before it can't.
Bots & Automation
Automated scripts and programmatic form-filling produce timing patterns that are fundamentally non-human. Caught from session 1 with no baseline required.
Unknown Impostors
Someone with stolen PII behaves differently from the real person — they paste data, switch windows, show read-type-check rhythm. Detected even without a prior baseline.
Familiar Deviation
A family member or insider who knows all the answers, types naturally, and passes every traditional check. They know the PII — but they can't replicate the behavioral signature.
KBA passes them. Document verification passes them. Device fingerprinting passes them. Population ML passes them. Stradum doesn't.
Flexibility
Form-Agnostic.
Device-Adaptive.
Any form, any field naming
Deploy on 10 different forms — applications, renewals, claims, account updates. The field mapping system normalizes everything. Behavioral signatures compare correctly across all of them.
Any device
Full signal set on desktop. Keystroke dynamics and navigation timing on mobile. The engine works with whatever signals are available and adjusts confidence accordingly.
Works with any framework
Dashboard
Monitor, Investigate, Act
A full analytics dashboard for behavioral intelligence — included with every plan.
Get Early Access
Start with a free 30-day pilot. We'll reach out within 24 hours to get you set up.
You're on the list!
We'll reach out within 24 hours to get your pilot started.
Close the Gap in Your Stack
One script tag. Zero friction. Behavioral signals no other tool can provide.