Device & Behavioral Intelligence — dEdge

Know who—and what—is behind every interaction

See beyond credentials and transaction data with native device and behavioral intelligence combining fingerprinting, risk assessment, behavior analytics, and edge computing across Android, iOS, and web. dEdge helps teams identify risky environments, AI agents, connected activity, and abnormal behavior early enough to change the decision.

Request a Demo →

100+ device signals across Android, iOS, and web

Fraud & Risk Platform/Device & Behavioral Intelligence — dEdge
Solution Pillars

Build a richer device risk profile

Evaluate device, browser, and behavioral signals together to see what static identity and transaction data alone cannot show.

See the device

Build a richer device risk profile

Evaluate device and browser attributes including user agent, browser type and plugins, current URL/referrer, OS and timezone, language settings, screen dimensions/resolution/color depth/pixel ratio, cookies and storage support, CPU and memory, graphics information, device type, and integrity indicators such as incognito mode, WebDriver, ad blocking, and suspicious browser or OS changes.

Device & browser fingerprinting · integrity indicators
See the behavior

Recognize when the interaction itself looks wrong

Use behavioral biometrics like typing speed, paste, autofill, mouse movement, click timing, page duration, navigation, and session-switching behavior to add context that static identity or transaction fields cannot provide alone.

Behavioral biometrics · session analytics
See automation and tampering

Surface environments designed to evade controls

Detect specific automation indicators like AI agents driving the browser, bot patterns, browser automation, WebDriver, emulator/cloud-phone, app-cloning, hooking, tampering, and other integrity indicators before they blend into downstream activity.

AI agent, bot & automation detection · anti-tampering
Connect activity across entities

Turn a device into network intelligence

Produce persistent device IDs that can be used with account and entity identifiers to link devices, accounts, identities, IPs, and other entities to reveal shared infrastructure and coordinated behavior across seemingly separate users and sessions.

Persistent device IDs · cross-entity linkage
New · AI Agent Detection

Know when an AI agent is driving the session

AI agents can now log in, fill forms, and move through a session the way a customer would. dEdge identifies when an agent is operating the browser, so you can decide what that agent is allowed to do before money moves.

Combine agent signals with device history, velocity, allowlists, and Knowledge Graph context to separate a customer's own agent from a fraudster's, then apply step-up, review, or block where risk warrants.

  • AI agent score: a 0–100 score for how strongly session behavior looks like an AI agent driving the browser rather than a person
  • AI agent vendor: the agent product recognized on the page
  • AI agent environment: a flag on every device token showing whether the page is running inside a known AI agent product
  • Validated against leading AI agent products, with human baselines across desktop and mobile browsers
  • Available today for web sessions through the dEdge JavaScript SDK
Product Proof

The signals behind the risk

A device profile, behavioral signal view, and linked-device graph make the intelligence tangible — from SDK/event collection, through device and behavior signals, to a risk score and downstream action.

SDK telemetry
→
Device ID
+ risk
→
Behavioral
signals
→
Rules · ML ·
Graph
→
Decision +
linkage
From SDK telemetry to a real-time decision: device ID and risk, behavioral signals, rules, ML, and graph context feed every downstream action.
How It Works

From session telemetry to a real-time risk signal

1

SDK collects device, browser, OS, hardware, network/IP, geolocation, integrity, and session-behavior data.

2

SDK initialization and token generation are followed by device-ID, device-risk, and behavioral-signal generation.

3

The customer backend sends the token to DataVisor; returned device and behavioral risk signals feed rules, models, and actioning.

4

Persistent device IDs and shared device/account context support linkage across sessions and accounts.

Device intelligence connected to the fraud platform

Give every signal more context

dEdge combines device fingerprinting, behavioral intelligence, risk assessment, and edge computing in one native layer for mobile and web. It becomes more powerful when evaluated alongside transaction history, identity signals, unsupervised ML, and Knowledge Graph relationships in the same risk decision.

Put It To Work

Apply device and behavior intelligence where identity can be manipulated

Strengthen specific use cases using linked devices, shared infrastructure, and coordinated behavior — catching risk earlier in the journey.

Account Opening Application Fraud Account Takeover Credential Stuffing Agentic Fraud Fraud Ring Detection Mule Detection
Proof & Outcomes

Prove the lift from richer context

Signal Coverage
100+
Device signals available across Android, iOS, and web

More than 100 device signals give risk teams the context to catch fraud earlier, reduce unnecessary friction, and strengthen every downstream decision.

Connected Platform

Turn device signals into decision intelligence

Feed dEdge directly into Rules, supervised and unsupervised ML, real-time decisioning, case review, and linkage/Knowledge Graph analysis.

Rules, Features & Workflow → Supervised Machine Learning → Unsupervised Machine Learning → Real-Time Decisioning → Case Management → Knowledge Graph & Link Analysis →
Built To Defend The Signal

Protect the integrity of client-side intelligence

dEdge is designed to keep device and behavioral intelligence trustworthy from collection through decisioning, even in adversarial environments.

  • Encrypted and signed tokens
  • HTTPS transport security
  • Key-based authentication or IP allowlisting
  • Per-device encryption keys
  • Anti-spoofing, anti-tampering, and obfuscation / reverse-engineering protection
FAQ

Common questions

What signals does dEdge collect?

dEdge analyzes more than 100 device, browser, network, integrity, automation, and behavioral signals to identify suspicious environments, connections, and activity that traditional identity checks can miss.

Can dEdge detect bots, emulators, or tampered environments?

dEdge includes supported indicators for automation and device integrity, including signals associated with emulators, webdriver activity, app cloning, hooking, rooting, and tampering.

Can dEdge detect AI agents?

Yes. For web sessions, dEdge identifies when an AI agent is operating the browser and returns three signals: an AI agent score, the agent vendor recognized on the page, and an AI agent environment flag on every device token. dEdge tells you an agent is present; evaluating those signals with device history, velocity, allowlists, and the action being taken helps separate a customer's own agent from a fraudster's.

How is dEdge used in a fraud decision?

dEdge risk signals can be evaluated alongside rules, machine learning, identity, transaction, and graph intelligence within DataVisor's real-time decisioning layer. To implement, integrate mobile SDKs for Android/iOS or a JavaScript SDK for web, and send the generated token from the customer backend to DataVisor.

See What A Transaction Alone Can't Tell You

Explore the device and behavior behind the event

Walk through a live risk scenario and see how dEdge surfaces device, behavioral, and relationship signals before the decision.

Request a Demo →