WoWyrd
Rule review

How your fraud rules perform, measured on your own data

WoWyrd measures how the fraud rules you already run perform on your own data, inside your own environment. It shows which rules to retire, merge, retune or add, with the effect in your own costs.

It works from files your current rule engine can export: your transactions, your rule-hit and decision log, fraud labels and your unit costs. It changes nothing in your engine. Available from release v0.4.0, during a pilot.

What you get

  • A technical report, rule by rule: hits, precision, overlap with other rules, decay over time, rules that never fire, and proposals with their evidence.
  • A short management report: headline metrics, charts, the top actions, and their value in your own unit costs.
  • If you share your rule file, WoWyrd re-runs your rules on your transactions and compares the hits with your own log. A rule is reviewed further only when the two agree.
  • A threshold search on your history: WoWyrd tries values around each threshold and proposes one only if it wins on both the earlier and the later part of your history. Each proposal, and each "keep", has a one-line reason.
  • For your AML rules, a sample of payments just under each amount threshold, for your analysts to label.
  • From the second report on, at least 29 days later: what changed since the previous report, and which requirements are still open.

Your data stays with you

  • The collector runs in your environment, on your files.
  • The reports stay in your environment. The summary export holds totals only: it hides counts under 10 and refuses to write if it finds an identifier. You decide what leaves.

Measured, with the caveats shown

  • Every proposal shows its evidence and its caveats, for example fraud labels that arrive late, and declined payments, which have no outcome.
  • Each effect is a backtest on your history, not a forecast. Nothing changes until your team approves it in your own engine.

Your rules, mapped to what your regulators require

We map your rules to what your regulators require, and flag when a new requirement affects them.

  • Each line runs from a requirement to the rule that answers it, and to that rule's metric, measured on your own data.
  • We follow the official publications of regulators in Ukraine, the EU, the UK and the US.
  • Ukraine: the National Bank's draft rules on the control of payment operations expect a monthly backtest review of every rule, two effectiveness metrics, written thresholds, and a decision to keep, adjust or cancel each rule. This is what the review measures. The text is still a draft.
  • United States: NYDFS Part 504 expects ongoing analysis of rules and thresholds, and the FFIEC BSA/AML manual expects documented threshold tuning.
  • European Union: the Payment Services Regulation is coming; it is agreed but not yet adopted.

This is not legal advice. Your legal and compliance teams decide what a requirement means for you.

How a pilot runs

  1. The collector runs in your environment, on your rule-hit and decision log.
  2. We go through the technical report and the management report with you.
  3. If your current tools are enough, we will tell you.

Ask how your rules perform on your own data.