← The platformAlerts & recovery

Failures, followed up.

Spend alerts reach customers before the invoice surprises them. Failed charges get a diagnosis — insufficient funds, expired card, bank block — and a recovery plan to match, not a blind retry loop.

Decline diagnosisSmart retriesDunning emails
Recovery · TokenfoldRecovered
Card declinedMon 11:40
insufficient_funds — not expired.
Retry scheduledThu 09:00
Payday-window retry + dunning email.
$1,240 collectedThu 09:02
Invoice settled automatically.
Why it matters

Recovery that reads the reason, not just retries.

01Diagnosis before action

Insufficient funds, expired card, bank block — each gets its own timing and message, not a one-size retry schedule.

02Up to 71% recovered

Diagnosis-first recovery, paired with the agent's autopilot mode, recovers well above a blind retry schedule.

03Every action logged

What triggered a retry, what was sent, and what it recovered — all visible in the audit log.

How it works

Three steps. No engineering ticket.

1A charge fails

The decline reason — insufficient funds, expired card, bank block — is captured immediately, not guessed at.

2Diagnosis picks the playbook

Each failure type gets its own retry timing and message; expired cards skip retries for an update-method email instead.

3Recovery runs automatically

On autopilot or approve-first, the agent executes the playbook and logs the outcome — no chasing email required.

71%of failed payments recovered, vs. ~30-40% on fixed retries
38 secmedian time from signal to action
100%of actions logged with the signal that triggered them

Send your first event today.