Device Fingerprinting
Every device gets a unique fingerprint the moment it touches your platform.
That identity survives cookie clears, incognito sessions, and IP rotations. Once a device is known, it stays known.
Product
Every device gets a unique fingerprint the moment it touches your platform.
That identity survives cookie clears, incognito sessions, and IP rotations. Once a device is known, it stays known.
Every session is scored from flags, history, behaviour, and location — not one check alone.
You set what score triggers what action. Keverd returns the number.
A persistent record across every session. The longer Keverd runs, the harder it is to hide suspicious behaviour behind a fresh login.
| Flag | What it means |
|---|---|
| BOT | Session behaviour matches non-human patterns |
| AUTOMATION | Form submitted by a script, not a human |
| USER_AGENT_SPOOFED | Device is misrepresenting its browser or OS |
| TIMEZONE_IP_MISMATCH | Device timezone does not match IP location — VPN or proxy use |
| AD_BLOCKER | Device is running an ad blocker |
One flag is context. Multiple flags together are a pattern — combined with your suspect score.
Surfaces when an unknown device appears inside an account — unusual device, location, or time — before the attacker finishes.
Device relationships tie independent-looking accounts to a single operation — whether it's fifty signups or hundreds.
Tracks device continuity. When a SIM swap breaks that link, you're alerted before OTPs or transfers go through.
Flags sessions where the OTP likely didn't come from the real account holder — before the transaction layer.
Flags simultaneous applications across platforms — upstream of the bureau, before disbursement.
Detects payments from devices or locations that don't match the account's normal behaviour.
Surfaces device clusters behind bonus farming — not two hundred users, twelve devices cycling registrations.
BOT and AUTOMATION flags fire at the session layer — before credential stuffing or bulk signups hit your system.
ALLOW
Low score, no flags, clean history. No interruption.
REVIEW
Score or flag crosses your threshold. Routed to a review queue — your team decides.
BLOCK
CRITICAL tier or BOT + AUTOMATION active. Request terminated. Device logged.
The Keverd response is returned before your system processes the request. Detection after the fact is forensics. Detection before the fact is protection.
Views tailored by industry — loan risk for lenders, member protection for SACCOs. Filter by risk tier, flag, date range, or device.
| Industry | Problem | Use case |
|---|---|---|
| Digital lenders | Catch hundreds of fraudulent applications — same device, different identities. | Application Fraud Prevention |
| Betting platforms | Bonus abuse prevention, multi-accounting detection, and account integrity. | Bonus Abuse Prevention |
| Ticketing companies | Scalper detection, bot blocking, and organiser account protection. | Scalper & Bot Blocking |
| SACCOs | Member account protection for digitizing savings & credit cooperatives. | Member Account Protection |
| Insurance platforms | Claims fraud detection, quote engine protection, and policyholder security. | Claims Fraud & Scrape Detection |
| Payment platforms | Credential stuffing defence, merchant protection, and payment fraud prevention. | Payment Fraud Prevention |
Your industry not listed? Custom templates are available on request.
Coming soon
Map connected devices across platforms — who coordinates with whom, and what they've done. One node flagged, the whole ring surfaced.
Coming soon
Opt-in, privacy-preserving shared threat intelligence. A device flagged yesterday on one platform is known when it appears elsewhere today.
Coming soon
Session logs, decisions, and flag history formatted for CBK, IRA, and SACCO frameworks — before the auditor asks.