Radar
Every fraud rule you write declines some good customers. Most tools never tell you how many. Pyncht Radar shows the revenue a rule blocks alongside the fraud it stops, so you can decide whether the trade is worth it.
Each rule reports blocked fraud and blocked good volume side by side. No rule ships without both numbers.
Risk scores trained on your customers and your decline history, not an industry average.
Backtest a rule against your last 90 days before it touches live traffic.
Ethoca and RDR alerts resolve a meaningful share of disputes before they become chargebacks.
Trigger step-up only where liability shift is worth the conversion cost, not on every transaction.
Evidence assembled from the charge, the session, and the fulfilment record, submitted on your behalf.
Clean, versioned, and idiomatic. The same call works whether Pyncht is analysing your existing processor or moving the money itself.
Read the API referenceblock if :risk_score: > 78
and :card_country: != :ip_country:
and :amount_in_usd: > 500
// backtest: blocks $12,400 fraud, $1,850 good volumeOnly if you tell it to. Default posture is to surface the trade-off, not to tighten it — most merchants find they are blocking too much, not too little.
Separately. Avoided dispute fees are a cost saving; the rescued transaction is recovered revenue. Your analysis never merges the two.
Connect a processor read-only and get a full cost breakdown. Nothing moves until you decide it should.