Fraud prevention and AML

Risk and compliance teams use Grappi to add network, mobile, document and watchlist signals to their decisions, and to screen customers as anti-money-laundering rules require.

How it fits together

Catch fake identities, mule accounts and sanctioned parties.

  1. Network and device

    Flag VPNs, proxies, unexpected countries and networks that do not fit the customer.

  2. Mobile and email

    Spot revoked, brand-new or mismatched numbers and disposable email addresses.

  3. Documents and faces

    Catch edited documents, photos of screens and synthetic faces.

  4. Watchlists and records

    Screen against sanctions, PEP, fraud and defaulter lists and court records.

  5. One decision

    Combine the signals into one score to approve, review or reject.

The checks

The 16 checks in this use case. Each one is switched on for your account separately, after your payment.

  • Fraud score, proxy, VPN and Tor flags, network and location for an IP address.

    • Flag sign-ups and logins from VPNs, proxies and Tor.
    • Add a fraud score to sign-up and payment decisions.
    • Spot bots and multiple accounts abusing offers and referrals.
  • Whether an IP address is inside the expected country and state.

    • Make sure customers sign up from India, or from the states you serve.
    • Block a service in regions where you cannot operate.
  • Checks that the network a customer connects from fits the rest of their identity.

    • Flag sign-ups whose network does not fit the stated location.
    • Spot account takeovers from unusual networks.
  • Whether a number is on the Department of Telecommunications revocation list, with its fraud risk indicator.

    • Stop sign-ups and payouts on numbers that have been revoked.
    • Flag numbers marked for financial fraud risk before lending or onboarding.
    • Re-check numbers on file before an OTP is sent for a sensitive action.
  • Checks a number against the telecom operator’s record without sending an OTP.

    • Verify a number where an OTP would interrupt the customer, such as at checkout.
    • Add a quiet second check on the number behind a high-value action.
  • Roughly how long a mobile number has been in use.

    • Treat brand-new numbers with more caution in fraud and credit checks.
    • Add the age of the number as a signal in risk scores.
  • Whether an email address exists and can receive mail, with disposable addresses flagged.

    • Block disposable and non-existent email addresses at sign-up.
    • Keep contact details clean for statements and notices.
  • Checks that an email address fits the rest of the customer’s details.

    • Flag sign-ups that use an email address unrelated to the applicant.
  • Scores document images for signs of tampering.

    • Flag edited ID and address documents before approval.
    • Review loan and insurance claim documents for tampering.
    • Send only suspicious documents to manual review.
  • Looks for signs that an ID image was edited.

    • Detect edited ID images before they reach approval.
    • Put suspicious uploads first in the manual review queue.
  • Checks from one selfie that a live person is in front of the camera.

    • Stop sign-ups that use a printed photo or a screen instead of a live person.
    • Add a liveness step to selfie onboarding without extra actions from the customer.
  • Whether a face image or video frame was generated or swapped by software.

    • Catch synthetic or swapped faces in selfie and video onboarding.
    • Protect video KYC from fake identities.
  • Screens a person or business against sanctions, PEP, defaulter and enforcement lists.

    • Screen customers as anti-money-laundering rules require.
    • Check directors, promoters and vendors before onboarding.
    • Re-screen existing customers on a regular schedule.
  • Checks a person against known fraud and defaulter lists.

    • Stop known fraudsters and defaulters at onboarding.
    • Screen customers before extending credit.
  • Searches court records for cases against a person.

    • Check borrowers, directors and senior hires for court cases.
    • Due diligence on business partners and vendors.
  • Combines several identity checks into one trust score.

    • Decide to approve, review or reject a customer from one score.
    • Auto-approve low-risk customers and cut manual review.

Start with the checks you need

Create an account, add money to your wallet, and subscribe to the APIs above.