Onboarding & Account Fraud
One device graph from signup to takeover
One device opening ten accounts is one device to us, even in private mode behind a new IP. The same graph that catches the duplicate signup recognizes the trusted returning customer, so friction lands on abusers instead of buyers.
90+
Device and runtime signals
One device
However many accounts
Real-time
Signup and login decisions
Evidence
Signals behind decisions
Scan
Collect signals in this browser and derive a device ID from them.
Wiping really clears local storage for this site in your browser, so sign-in state on this site may be affected. Cookies are not touched.
Wipe
Clear localStorage and sessionStorage on this origin, for real.
Re-scan
Derive the ID again, from signals alone, with storage empty.
run the scan to list the real keys stored here
This demo derives the ID from measured signals client side, in your browser, so you can watch it happen. Production matching runs server side, with fuzzy matching over the collected signals rather than a single exact hash.
The Problem
Identity checks stop at the document
KYC verifies a person exists. It does not tell you the same laptop has opened nine accounts this week, or that tonight's login to a good account is not coming from its usual device. Emails, phone numbers, and IPs are all cheaper to rotate than hardware.
The Solution
Recognize the device, then decide
LRDefender anchors identity below the storage layer, so clearing cookies, going incognito, or switching browsers within the same engine family does not reset an abuse budget. Network verdicts return their evidence labelled fact or inference; device and automation decisions return their contributing signals, so an analyst can defend a decision instead of quoting a score.
Duplicate detection at registration
Flag a new signup that shares hardware and environment signals with existing accounts before the welcome email sends, and before the bonus pays.
Login risk with explainable signals
Every attempt returns a decision plus the contributing factors, new environment, automation hints, mismatched device, so analysts resolve queues in minutes instead of guessing.
Session integrity on sensitive actions
Call identify() again before password changes, payouts, or API token issuance and compare the device against the account's history, which catches token theft that a valid session cookie hides.
Account linking in the graph
See when many accounts share a small set of devices, or one device rotates through many identities, with the linkage evidence attached to each edge.
Promo and referral guardrails
Key your promo codes and invite credits to the device's history instead of an email address, your promo system on our device ID, so star-shaped referral farming stands out from organic sharing.
Proportional responses
Allow, challenge, flag, or block, with stricter thresholds on money movement and lighter ones on browsing, your call at each step, all against the same device graph.
How It Works
Three steps to protection
Instrument the moments that matter
Add the Browser SDK to signup, login, coupon apply, and payout flows, the places where abusers reveal themselves.
Decide with the graph
One API call returns the device link, account-linkage signals, how many accounts this device has touched and how many devices share this account, and the automation verdict, in real time on the request path.
Review with receipts
High-risk decisions carry a device timeline and their contributing signals, so a human review takes minutes and a wrong block gets overturned fast.
Fund growth, not repeat abusers
Run LRDefender beside your next campaign or on your login path and review the linked accounts with us. Your thresholds, your rules, evidence for every decision.