Skip to main content Overview The Feedzai connector enables financial institutions to collect and assess bank customers’ behavioral data, allowing them to detect potential fraudulent activities and improve their risk management strategies.
This connector is available on Backbase Marketplace and integrates with Backbase Engagement Banking Platform out of the box. You can obtain a Feedzai contract directly through Backbase for faster time-to-market.
The connector uses the Feedzai API suite (v5) and its AI and machine learning capabilities to assess each user interaction in web or mobile apps and instantly generate a corresponding risk score. This risk score allows financial institutions to trigger the appropriate Level of Assurance (LOA).
Key Features
AI and machine learning : Real-time risk assessment using advanced algorithmsBehavioral data collection : Capture user interaction patterns and browsing behaviorDevice intelligence : Collect device information, location data, and VPN usageReal-time risk scoring : Instant risk score generation for every user sessionRisk-based decisioning : Configurable responses based on risk levelsComprehensive analysis : Combine behavioral and device data for accurate fraud detection
How It Works Assess User Activity Risk When a bank customer uses their web or mobile app:
Data Collection : Device profiling code collects behavioral and device intelligence data
Behavioral data: Browsing speed, interaction patterns, left/right-handed behavior
Device intelligence: Device type, location, VPN usage, known device status
Risk Engine Analysis : The app generates a unique risk engine ID and makes an API call through the Grand Central Unified API Specification
Risk Engine evaluates all collected data in real-time
Applies predefined rules to assess risk dynamically
Generates comprehensive risk score with decision reasoning
Risk Assessment : API response contains:
Risk score based on extensive user activity analysis
Reasons for the scoring decision
Recommended Level of Assurance (LOA)
Level of Assurance (LOA) Categories The Feedzai connector provides the following risk score categories:
LOA Low : Financial institution can allow user to proceed with activityLOA Substantial : Financial institution can enable additional authorization and authenticationLOA High : Recommended to enable additional authorization and authentication before proceedingBlock : High risk - recommended to stop user from proceeding with activity
Risk score thresholds are configurable according to the financial institution’s risk appetite.
Detection Capabilities Behavioral Analysis
Bot detection : Identify botnets and Remote Access Trojans (RATs)Pattern recognition : Detect unusual browsing speed or interaction patternsBehavior consistency : Compare current behavior with historical patternsHuman verification : Distinguish between human and automated interactions
Device Intelligence
Device recognition : Identify unknown or suspicious devicesLocation monitoring : Detect unusual locations or geographic anomaliesVPN detection : Identify use of VPNs or anonymization toolsDevice fingerprinting : Track device characteristics and changes
Use Cases Payment Fraud Detection Real-time scoring of payment transactions to identify potentially fraudulent activity before execution.
Account Takeover Prevention Detect unauthorized access attempts by analyzing deviations from normal user behavior and device patterns.
Transaction Monitoring Continuous monitoring of all user transactions with automated risk assessment and response triggering.
Step-Up Authentication Trigger additional authentication requirements based on risk score thresholds for high-risk activities.
Session Termination Automatically end user sessions when risk scores indicate high probability of fraudulent activity.
Configuration Configure your Feedzai connector through the Grand Central platform:
API endpoints and integration parameters
Risk score thresholds for each LOA level
Automated response configurations
Device profiling settings
Event routing and notifications
Supported Operations Through the Grand Central Unified API Specification:
Initiate risk assessment for user sessions
Retrieve real-time risk scores and analysis
Access detailed reasoning for risk decisions
Configure risk thresholds and responses
Monitor ongoing user sessions
Query historical risk assessments
Integration Benefits
Real-time protection : Instant fraud detection during user sessionsComprehensive analysis : Multi-factor risk assessment combining behavioral and device dataConfigurable responses : Adapt risk thresholds and actions to your risk appetiteAutomated decisioning : Reduce manual review with AI-powered risk assessmentEnhanced security : Strengthen fraud prevention across all digital channelsImproved user experience : Seamless protection without impacting legitimate usersCompliance support : Meet regulatory requirements for fraud prevention and AMLScalable solution : Handle high-volume transaction analysis with AI efficiencyBIAN compliance : Standardized security interactions based on BIAN framework 