Workspace/Overview
Soft declines
0
Pending
0
Scheduled
0

Only scheduled retries reach the engine.

Recovered
0%
0
Soft0Hard0
Failed
0%
0
Soft0Hard0
Lost
0%
0
Soft0Hard0
Recovery ML
Adaptive decisioning
Idle
Idle·0 processed·0 in queue
Retry budget
1 / 3
Recovery rate0%
Processed0
adaptive scoring · 30-day forecast window

When ML retries

Pattern recognition

ML clusters similar failures by parameters like time of day, region, and issuer — then triggers retries during proven peak-success windows.

Error type · select to inspect pattern
All failuresretry success rate · last 30 days
ML retried here
ML waited
30-day mean
0%25%50%75%100%Day 1Day 5Day 10Day 15Day 20Day 25Day 30
Best42%· No standout window — peaks cancel out
Worst34%· Mediocre across the board

Real-time simulation

Sandbox

Configure a failure scenario. Watch ML decide whether to retry, when, and how — recover or write off.

Failure class
Error type
Day of week
Hour of day · 3p
12a6a12p6p11p
Region
Issuer
Invoice amount · $480
FAIL-XXXXXXXX
$480 · Insufficient funds · Mon 3p · Chase · NA
SOFT
ML pre-decision
Best window: Fri 10a · expected 85%
Retry now would land at 43% — ML would wait 115h.
Timeline
Click Run simulation to start