Does the client need to blink or turn their head?

No. Passive liveness needs one selfie — no head turns and no command sequence. An honest investor does less and continues account opening faster.

What happens if someone uses a photo, video or deepfake?

Liveness checks whether a live person is in front of the camera. Face match with the document, device analysis and IP analytics add to the result.

Does biometric.vision make the decline decision?

No. The platform returns check results; approve, decline and manual-review rules are set by the broker for its risk profile and internal procedures.

Can the client be re-confirmed without a full KYC?

Yes. After the first check the investor’s face can be matched to the saved template. For sensitive actions or higher risk the broker decides which steps to repeat.

How does the solution support the AML process?

Sanctions and PEP screening can be added as a separate stage. Confirming the investor’s identity stays at the core; AML checks complement the result in the compliance process.

Where is biometric data stored?

There is no universal answer: it depends on architecture, chosen stack and contract terms. On a demo the team will walk through your scenario and requirements for storage, processing and deletion — without promises that may not fit your jurisdiction.

Biometrics for brokers

Open accounts for investors — not for nominees

One scenario — one verdict: document, passive liveness and face match work together, and the decision fits into the broker’s rules. The investor needs one selfie, with no series of camera commands. They open an account from a smartphone, while someone else’s passport or a deepfake is not a pass to trade.

Start free Request a demo

Session · broker_onboardingSuccess

matchedface ↔ document

Document captureddata extracted

Liveness checkpassed

Video spoof analysisno signs found

Device signalsin verdict

Session context · runtime, network, model and OS

Account can be openedVerdict returns to your onboarding flow

Sound familiar?

The document is real. The investor is not.

The form is filled in, the passport passed a first check. But the person in front of the camera may not be who the application claims: someone with another person’s document, a deepfake — or a stand-in who opens an account with their own real passport, but not for themselves.

If the truth comes out after the account is open, compliance has to clean it up. Make the path heavier and you lose some honest investors on the way. Check only the document — and you still do not answer who is opening the account right now.

Registration streamlive

Application 4831Deepfake signs on camerarisk

Application 4832Someone else’s document · nomineerisk

Application 4833Suspicious session environmentreview

Application 4834Document · liveness · face · deviceapprove

Investors and schemes walk through one door — split them before the account opens.

How verification works

Investor identity is confirmed in one scenario

For the client everything is sequential and without extra steps. The broker gets a single result and decides what to do next: open the account, stop the application or send it to manual review.

Document

The investor photographs a passport or ID. The platform recognizes the data and checks the document — then the next step can begin.

Step 1 · documentCaptured

Document typerecognized

Document dataextracted

Portrait from documentready for match

Document in sessionData available for the next verification layer

One selfie and face match

One frame is enough for passive liveness — no head turns. The selfie face is matched to the photo in the document. Match accuracy is 99.9997%.

Step 2 · liveness and faceSuccess

Selfie

Document

Live person on cameraconfirmed

Face match with documentmatched

Face confirmedMatch threshold configurable to your risk rules

Spoof analysis

Photo, replay, mask or deepfake are analyzed separately from face-to-document match. The result enters the verdict before the account opens — as a risk signal for your rules, not an automatic sentence.

Step 3 · face and video spoofLayer passed

Photo instead of facenot found

Recorded videonot found

Masknot found

Deepfake signsnot found

Video stream substitutionnot found

Layer result — in verdictSeparate from face-to-document match

Device and IP signals

Device analysis and IP analytics are added to the verdict. A single suspicious signal alone does not mean an automatic decline — it is context next to document, liveness and face.

Step 4 · Device IntelligenceContext collected

Simulatorno

Emulatorno

Rootno

VPNcontext

Proxy / TORno

Model and OSknown

VPN — context, not a verdict — evaluated with the other layers

Session signals addedSame flow, no separate anti-fraud form

Verdict

Document, liveness and face results return to your account-opening flow. Approve, stop or manual review — by the broker’s rules.

Step 5 · session verdictReady

OverviewLivenessFace matchEvents

14:02Document captured

14:02Liveness confirmedok

14:03Face matchmatched

14:03Video spoofnot found

14:03Device signals accounted for

Approverule A

Reviewrule B

Declinerule C

Responseto your flow

You make the decisionWe return layer results and final status

Application context

The selfie passed. What about the device and network?

A face match is not the full application context. Device Intelligence analyzes the device, and IP Check adds IP analytics. The risk team sees this before the client gets access to trading.

A single suspicious signal alone does not mean an automatic decline. It is a reason to look closer at the application, ask for an extra check or send the case to manual review.

Overview · Device IntelligenceRisk

Face passed checksDocument, liveness and match — no issues

Emulatordetected

Rootdetected

VPN / proxycontext

TORcontext

Device modelknown

OSknown

!Registration — additional reviewAction is defined by your risk rules

After the account opens

A known investor gets a short entry. A duplicate gets a new check

At account opening the client already confirmed identity. On a return login or a sensitive action they can be matched by face to the saved template — without a full document scenario. A chain of accounts on one face stops being invisible: the applicant’s face is compared to templates already saved. Device and IP analysis add context. Average face-database search speed is about 1.5 seconds.

Face Search · 1:NSearch complete

New registrationFace compared against the connected database1:N

match

Average face database search speed~1.5 sec

!Potential face repeatSignal enters the verdict before trading access

For developers

Classic integration

If your team prefers to assemble the architecture itself, connect ready SDKs and APIs.

JS · Webview

Web SDK / Flow Widget

Ready JS camera component for web registration or in-app Webview. Verification runs inside your onboarding; design adapts to your brand.

JavaScriptCustom designWebview

iOS · Android

Mobile SDK

Native verification flow in mobile apps: camera access and biometric micro-packets even on weak connections.

SwiftKotlinOffline-ready

Backend

REST API

Layer results and final verdict arrive in your backend logic: risk rules and decisions stay on your side. X-API-KEY auth.

RESTX-API-KEYServer-side

AI · AGENTS

MCP server for AI agents

Connect verification to agent scenarios without writing code. Works with Claude, GPT and any MCP client.

MCPClaudeAI-agents

broker_onboarding — REST API

# Client onboarding / broker_onboarding

POST /v1/verification/check/

Host: api.biometric.vision

X-API-KEY: sk_live_••••••••••••••

Content-Type: application/json

{ "scenario": "broker_onboarding", "checks": ["document", "liveness", "face_match", "device"] }

# → 200 OK

{ "status": "completed", "live": true, "face_match": "match", "device": { "vpn": true } }

99.9997%match accuracy

~1.5 secface database search

1 selfiepassive liveness

DI + IPsession context

Questions and answers

What brokers usually ask

A document alone is not enough here. The applicant takes a selfie, passes passive liveness, and their face is matched to the document holder’s photo. The broker gets all results in one session.

500 checks a month free

Check who is really opening the account

Run the scenario in your onboarding: document, one selfie, face match and application context. See the honest investor’s path and the result the risk team gets.

Start free Request a demo

Numbers and facts

433+: Document types —Passports and hundreds of other document types worldwide
180+: Supported countries —Global coverage for verification across jurisdictions
40 sec: Onboarding time —User registration and verification in under a minute
99%: Technology accuracy —High-precision AI document recognition and verification

Open accounts for investors — not for nominees

The document is real. The investor is not.

Investor identity is confirmed in one scenario

Document

One selfie and face match

Spoof analysis

Device and IP signals

Verdict

The selfie passed. What about the device and network?

A known investor gets a short entry. A duplicate gets a new check

Classic integration

Web SDK / Flow Widget

Mobile SDK

REST API

MCP server for AI agents

What brokers usually ask

How does verification stop a nominee with someone else’s passport?

Does the client need to blink or turn their head?

What happens if someone uses a photo, video or deepfake?

Does biometric.vision make the decline decision?

Can the client be re-confirmed without a full KYC?

How does the solution support the AML process?

Where is biometric data stored?

Check who is really opening the account