Introducing Kairos: real-time fraud decisions, from sign-up to payout.Read the launch post
Kairos
Account opening

Let real customers in. Keep fake accounts out.

Decide every sign-up as it happens. Kairos sees the devices, emails and addresses a new account shares with others, and how the sign-up session went.

SessionsAccount opening
  1. +0:00sign_up
  2. +1:37verification_started
  3. +2:14email_verified
  4. +3:51login
Review
  • shared_device0.94
  • rapid_attempts0.81
  • disposable_email0.62

Events you send

  • sign_up
  • verification_started
  • email_verified
  • login

Example signals

Signals are questions you write. Grayson answers each with a probability, and your models and policies use the answers.

  • shared_deviceThe device was used to open other accounts recently.
  • rapid_attemptsMany sign-up attempts came from this device or network in a short time.
  • disposable_emailThe email address is from a throwaway provider.
  • country_mismatchThe address country differs from the device and IP country.

What each outcome does

ApproveOpen the account.
Step upAsk for a document or a selfie.
ReviewHold for an analyst.
BlockRefuse the sign-up.

How it decides

  1. 1Your product sends eventsSign-ups, logins, orders, payouts and chargebacks, from the SDK, the API or webhooks.
  2. 2Kairos builds the contextIt links the request to its user, devices, cards and addresses, and to their history.
  3. 3Models and rules decideA workflow asks Grayson, scores with your model and applies your policies.
  4. 4You get an answerThe outcome, a risk score from 0 to 100, and the reasons, in one response.
WorkflowsPayout decisionsVersion 6
TriggerDecides withdrawal
Ask a decision modelGrayson · 6 questions
Score with a modelLogistic regression v6
If / elseScore is 80 or more
Create a caseHold for review
ReturnThe model decides

See Kairos on your own decisions.

Talk to us about your fraud and risk decisions, or start on Kairos Cloud today.

Try Kairos Cloud