Skip to main content
Version: 1.2

Integration Phases: Your Path to a Successful Integration

This guide outlines how onboarding to Fraudio works. Work runs across two parallel tracks:

  • Technical track: API integration, field mapping, data quality, historical data, and production readiness.
  • Risk track: use cases, rules, thresholds, risk appetite, shadow testing, validation, and tuning.
PhaseFocusWhat “done” looks like
KickoffAlign scope and set up accessPortal access, communication channel, and project requirements confirmed
Technical IntegrationBuild the core real-time integrationCore endpoints and required fields validated in production; initial rules configured
TuningComplete the integration and refine controlsSupplementary data in place where agreed; risk appetite and controls finalized
BAU / Ongoing OperationsSteady-state supportPerformance reviewed; handover to normal supported operations

Kickoff​

Establish the project foundation so both tracks can start execution.

Technical track​

  • Confirm technical contacts and the intended transaction flows.
  • Get portal access: invite users, assign roles, and create API keys. Use sandbox keys for early testing.
  • Review the API documentation and agree the initial endpoints, fields, and real-time actions for your scope.

Risk track​

  • Confirm the primary risk contact.
  • Share existing fraud/AML controls, compliance obligations, and initial priorities or use cases where available.
  • Align on intended actions for those use cases (for example review, deny, or monitor).

✅ Phase complete when: project requirements and setup are agreed, portal access works, and both tracks know what to deliver next.


Technical Integration​

Build the core real-time integration and put the first controls in place against your traffic.

Risk scoring status: During this phase, treat risk recommendations in the API response as unreliable. Data quality is still being established, so do not use scores or recommendations for operational decisions yet.

Technical track​

  1. Follow the Quickstart Guide
    Use the Quick Start Guide to create test API keys and send your first API call.

  2. Map your data to Fraudio's schema
    Map your internal fields to Fraudio's expected format. See the Mapping Guide.

    Focus first on the elements needed for production traffic on the core flow:

EndpointApplicable ProductsApplicable Integration Types
Payment Fraud ScorePFD, MIF, AMLPre-authorization, Post-authorization
Post-Authorization EnrichmentPFD, MIF, AMLPre-authorization
  1. Data quality review
    Fraudio assesses completeness, accuracy, and consistency so the data meets the minimum bar for production use.

  2. Start sending production traffic
    After data quality is approved, switch to production API keys and send live transactions through the agreed core endpoints. This is the point from which Fraudio receives your real production volume on the core integration.

Risk track​

  • Configure the initial agreed fraud/AML rules and workflows.
  • Define intended actions and priorities for the key use cases.
  • Where agreed, run initial rules in shadow so you can review results without affecting decisions.

✅ Phase complete when:

  • Payment Fraud Score and Post-Authorization Enrichment (where applicable) are sending production traffic with the required fields.
  • Core data quality checks have passed.
  • The first agreed rules are configured and, where agreed, producing results you can review (for example in shadow).

Go live​

Go live means the core integration is live in PRD: Fraud Score and Post-Authorization Enrichment (where applicable) are sending production traffic with the required fields, data quality is approved, and the first agreed rules are configured (in shadow where agreed).

Tuning starts after this.


Tuning​

With production traffic flowing on the core integration, complete any remaining data work and refine controls using observed behaviour.

Risk scoring status: Recommendations follow the rules enabled in the portal. Models start conservative and adapt as you tune thresholds and risk appetite with Fraudio.

Technical track​

  • Add agreed supplementary fields to the production payloads. See Important & Supplementary fields in the Fraudio Data Schema.
  • Implement remaining flows where they apply to your scope:
EndpointApplicable ProductsApplicable Integration Types
Payment Events EnrichmentPFD, MIF, AMLPre-authorization, Post-authorization
Dispute EventsPFDPre-authorization, Post-authorization
  • Provide historical transaction data where agreed (typically at least 24 months when available). A short sample first helps validate format before the full set. See the Batch Transfer Guide.
  • Configure monitoring-alert webhooks if required. See the Notification System.
  • Resolve remaining data-quality gaps and confirm integration stability.

Risk track​

  • Review performance on real production traffic: alert volumes, outcomes, and operational workload.
  • Refine risk appetite and thresholds.
  • Enable additional rules relevant to your strategy; move validated controls from shadow to enforced where appropriate.
  • Sign off the initial production configuration with Fraudio.

✅ Phase complete when: supplementary data agreed for the integration is available, historical data is validated where in scope, performance has been reviewed, and both teams are comfortable with the production configuration.


BAU / Ongoing Operations​

Onboarding is complete. You operate under Fraudio's normal supported model.

Risk scoring status: Rules and models are operational and adapted to your data.

Technical track​

  • Monitor integration health and maintain data quality.
  • Support enhancements: additional endpoints, fields, or products as your use case evolves.

Risk track​

  • Continuously review performance and tune rules and models as needed.
  • Add new use cases over time.
  • Keep an open feedback loop via the portal or support channels.
  • Optional scheduled reviews with Fraudio to align on product updates and optimisation opportunities.