Manual Review in KYC: Human-in-the-Loop Verification

The Importance of Manual ReviewManual review is essential for handling complex or ambiguous cases that automated systems can't resolve, ensuring accuracy and reducing false positives.

When to Use Human ReviewManual review should be triggered by specific warning signals, risk factors, or inconsistencies detected during automated identity verification processes.

Key Steps in Manual ReviewEffective manual review involves assessing warnings, inspecting documents, verifying biometrics, and analyzing risk signals to make informed decisions.

Didit's SolutionDidit's platform is designed to streamline manual review, providing a comprehensive dashboard with all necessary data and tools for efficient human-in-the-loop verification, including AML screening and ID verification.

Understanding the Role of Manual Review in KYC

In today's digital landscape, Know Your Customer (KYC) processes are crucial for preventing fraud and ensuring regulatory compliance. While automation has significantly improved the efficiency of identity verification, it's not foolproof. Manual review, also known as human-in-the-loop verification, remains a vital component for handling complex, ambiguous, or high-risk cases that automated systems can't resolve alone. Manual review ensures accuracy, reduces false positives, and enhances the overall reliability of your KYC process.

When to Trigger Manual Review

Knowing when to escalate a verification to manual review is critical. Here are several scenarios where human intervention is necessary: * Warning Signals: Didit's system flags verifications with warnings or inconsistencies. These warnings are specific signals that trigger the need for manual review. Examples include low liveness scores from Didit's Passive & Active Liveness detection, potential AML matches identified by Didit's AML Screening & Monitoring, or document inconsistencies detected by Didit's ID Verification. * Low Confidence Scores: When automated checks return low confidence scores (e.g., in face matching or document verification), a human reviewer can assess the evidence and make a judgment call. * Edge Cases: These are situations where the automated system is uncertain due to poor image quality, unusual data patterns, or other factors that require human judgment. * High-Risk Scenarios: Transactions or users that meet certain high-risk criteria (e.g., large transaction amounts, connections to sanctioned countries) should always undergo manual review. For example, if Didit's ID Verification detects a potential data inconsistency between fields on an ID document, the session moves to "In Review" status, requiring a trained reviewer to visually confirm the document and extracted data. ## Conducting an Effective Manual Review: A Step-by-Step Guide An effective manual review process involves several key steps. Using Didit's console, reviewers can follow these steps to make informed decisions: 1. Initial Assessment: Start by reviewing the high-level information. Check all warnings displayed in the session overview, such as low liveness score, AML match, or document inconsistency. Review the user's previous verification attempts by accessing the session history to identify any patterns or potential fraud attempts. Examine the session events timeline for a chronological log of every action taken during the session. 2. Document Inspection: If warnings are ambiguous, visually inspect the document images directly, looking for signs of digital editing, cropping, or physical manipulation. Check the extracted OCR data (name, date of birth, document number, expiry date) and confirm it looks consistent and plausible. Review document matches to identify whether the user has been verified before under a different identity or if the document appears in your blocklist. 3. Biometric & Liveness Verification: Review the face match score and warnings. Visually compare the selfie and the document portrait, focusing on structural features. Review the liveness detection score, keeping in mind that poor camera quality or low lighting can sometimes produce lower scores for genuine users. Check face search results for duplicate or blocklist matches. 4. Risk & Signal Review: Assess contextual risk by navigating through the AML, IP Analysis, and Database Validation sections. For AML screening, examine each matched entry in detail to determine if it's a true positive or a false positive. Check if the user is connecting through a VPN or proxy server. Review the device type, operating system, and browser information for anomalies. ## Approval, Decline, or Resubmission: Making the Right Call After completing the review, you have three options: Approve, Decline, or Request Resubmission. * Approve: Approve the session when the identity document appears authentic and unaltered, all data fields are consistent and legible, the selfie biometric match confirms the same person as on the document, the liveness check score is within acceptable range, no relevant AML matches were found, and device and location data are consistent and non-suspicious. * Decline: Decline the session when the document appears tampered, forged, or digitally altered, the selfie does not match the document photo, a high-confidence AML match is confirmed, there are strong indicators of a presentation attack, fraudulent identity patterns are detected, or critical device or location inconsistencies indicate a fraud attempt. * Request Resubmission: When a session cannot be clearly approved or declined due to fixable issues, request the user to resubmit specific verification steps. Use resubmission when the issue is likely correctable by the user, such as blurry document images, failed liveness check due to technical issues, wrong document submitted, or incomplete steps. With Didit, you can select the specific features to resubmit and optionally send an email notification explaining what needs to be redone. ## Best Practices for Efficient Manual Review To ensure your manual review process is as efficient and effective as possible, consider these best practices: * Establish Clear Guidelines: Create detailed guidelines for reviewers to follow, ensuring consistency and accuracy in decision-making. * Provide Training: Ensure your review team is well-trained on fraud detection techniques, document verification, and compliance requirements. * Document Everything: Keep a detailed record of each review, including the reasons for the decision and any supporting evidence. * Regularly Audit Your Process: Periodically review your manual review process to identify areas for improvement and ensure it remains effective. * Leverage Technology: Use tools like Didit to streamline the review process, providing reviewers with all the information they need in one place. ## How Didit Helps Didit's AI-native identity platform is designed to streamline and enhance the manual review process, offering a comprehensive suite of tools to aid human reviewers. With Didit, you benefit from: * Centralized Dashboard: Access all necessary data in one place, including document images, extracted data, biometric scores, and risk assessments. The verification table view lists all sessions with their current status, document type, country, and other key details at a glance. Verifications marked as In Review have triggered one or more warning signals during automated processing, making it easy to identify sessions that require your attention. * Automated Warnings and Flags: The system automatically flags suspicious activity and inconsistencies, ensuring reviewers focus on the most critical cases. * Comprehensive Verification Tools: Utilize Didit's ID Verification (OCR, MRZ, barcodes), Passive & Active Liveness detection, 1:1 Face Match & Face Search, and AML Screening & Monitoring to gather all necessary information for informed decision-making. * Resubmission Feature: Request users to resubmit specific verification steps when issues are fixable, improving completion rates and reducing drop-off. * Free Core KYC: Start with Didit's free tier, allowing you to implement robust KYC processes without upfront costs. Didit's modular architecture allows you to plug-and-play identity checks, orchestrate workflows with a no-code engine, and leverage developer-first tools with instant sandbox access and clean APIs. This makes Didit the ideal solution for automating trust and ensuring compliance.

Ready to Get Started?

Ready to see Didit in action? Get a free demo today.

Start verifying identities for free with Didit's free tier.