MEET FINGERLYA clearer picture of every session.Explore the signals
Interactive demo · no sign-up

One request. Everything it tells you.

Explore a session the way your server receives it: the visitor ID, the suspect score and every check behind it, and the identify response as the API returns it.

  • Nothing read from youThe session on this page is synthetic.
  • The real response shapeEvery field the identify endpoint returns.
  • Your decisionA score to act on, never a verdict.

One visitor. Every check.

The visitor ID belongs to the device and stays put; the request ID is new on every call. The tabs read this one request through six checks, and the JSON switch shows the response body.

app.fingerly.io / identification / events
PREVIEW

Session inspector

One synthetic session. Identify again, or open a check.

Visitor IDRETURNING VISITOR
X9pL2mRc7KvT4bQw8NdF

Recognised from the device itself, not from cookies or storage, so clearing them or switching networks does not reset it.

Confidence
100
Identifiable
Yes
Request ID · new on every call
01a0a84b-e6a2-7c09-9f51-0b3d7a26c8e4
Suspect scoreHIGH
557 signals4 of 6 checks flagged

The overall score for this request: the weight of every signal that fired, added up.

Device farm detectionFLAGGED

Is it one of many?

Detected

The “phone” is an emulator, and it shares its network with far more devices than your traffic ever puts in one place.

+14
Emulated device
Confidence Medium
+7
Device farm
Confidence High
Share of the suspect score+21 of 55
Evidence
Claimed device
Android 14 phone
Emulated
Yes
Devices on network · 1 hour
61
Devices on network · 24 hours
173

A signal is a clue, not a verdict. Your application stays in control of the decision.

Three layers. One answer.

Identification, the signals, and a score to read them by, on every request, from one call.

Layer 01 Identification

Which device is this?

A visitor ID for the device, not the browser’s storage, so clearing cookies, opening a private window or changing networks does not reset it.

  • Visitor ID
  • Confidence, 0 to 100
  • New or returning
How it works
Layer 02 Signals

What is it doing?

Every request is checked for the things that change how much to trust it, and each finding comes with its own confidence.

  • Bots and automation
  • VPNs, proxies and Tor
  • Emulators and device farms
  • Tampering and private windows
Every signal
Layer 03 Suspect score

How much should it worry you?

One number: the weight of everything that fired, added up, and a level to read it by. What each level means in your product is yours to decide.

  • A score with every trigger behind it
  • Low, medium or high
  • Your rules, your decision
Build a policy

Same signals. Different decisions.

Four moments where a product reads an identification. The rules are an example application’s: the score informs a decision, it never makes one.

01

The fifth free trial

A fresh email and a cleared browser, from a device that has signed up before, and is not the phone it claims to be. The trial is refused before it is granted.

Rule score ≥ 30 → refuseNew account fraud
02

Card after card

A person, on Tor, in a disguised browser, trying to pay far more often than a shopper does. Nothing on its own proves fraud; together it is past the line, so the order waits for a look.

Rule score ≥ 30 → hold for reviewPayment fraud
03

The right password, somewhere new

A device the account has never used, behind a VPN, in a private window. Medium, not high: enough to ask for a code, not enough to lock a customer out.

Rule score ≥ 15 → ask for a codeAccount takeover
04

A familiar laptop

The device this customer always uses, at home, at a person’s pace. Nothing fired, so nothing gets in the way: no code, no captcha, no second guess.

Rule score < 15 → let them inReturning user experience
[ IDENTIFY ][ UNDERSTAND ][ DECIDE ][ FINGERLY ]
Less guessing. More knowing.

Make the next connection a trusted one.

Start with the signals, keep your own decisions, and pay only for what you identify. No credit card needed.