Watches your stack.
Files your tickets.
Recovers your credits.
A vigilant AI-ops agent for lean SaaS teams. When Stripe breaks, AWS burns, Datadog fires or your deploy regresses, Recoupfox files the right ticket with full context — and chases service credits back into your books.
- First-touch
- under 90 s
- Recovery
- filed to paid
- Coverage
- 24 / 7 · Slack-native
Datadog · checkout-error-rate
Fired12.4 % over 5 m — p99 latency 1.8 s · 412 user-affected sessions
Filed · STR-3091 to Stripe Support
Attaching: charge log, affected customer list, attempted retry timeline · SLA window Sep 12 14:00 UTC
Credit negotiated · AWS Support
+ $3,420us-east-1 RDS outage 09:41–11:08 UTC · credit policy §4.2 applied · reflected in Sep invoice
What we watch
Four sources. One queue. No new dashboard.
Recoupfox sits on the alerts and webhooks your vendors already publish — billing, infra, observability and CI. One triage lane, Slack-native.
Stripe
Billing regressions, failed charges, refund spikes, subscription drift.
AWS
Usage spikes, untagged spend, idle reservations, service health.
Datadog
Monitor firings, SLO breaches, error-budget burn, anomaly alerts.
GitHub
Failing deploys, broken workflows, regressions in downstream CI.
How it works
Connects, watches, files. You stay in charge.
Three steps from OAuth to your first filed ticket — and every step is reversible from your side.
- 01
Connect your stack
OAuth in Stripe, AWS, Datadog and GitHub in under five minutes. Read-only scopes, scoped per environment, encrypted at rest.
- 02
We watch the signal
A quiet agent sits on every alert channel, anomaly dashboard and billing web-hook your vendors already publish — no new dashboard to babysit.
- 03
It files, you decide
Recoupfox drafts the right ticket with incident context attached, queues it to Slack for sign-off, and stays on the case until your vendor credits are back in your books.
What recovery looks like
Outages get fixed. Recoupfox makes them pay back.
Every quarter, Recoupfox closes the loop: filed tickets, vendor SLAs applied, credits back in the books. Here is one representative cohort.
- Service credits recovered
- $84,210
- Tickets filed autonomously
- 146
- Median first-touch on a fired alert
- < 90 s
across 11 vendor disputes last quarter
every one with billing + incident context attached
before the on-call rotation even pages
Q3 recovery ledger
Draft · reconciled Mon- AWSSep 09us-east-1 RDS incident+$3,420
- StripeSep 12webhook delivery slip+$1,180
- DatadogSep 17metrics ingest outage+$640
- GitHubSep 22Actions queue stall+$210
Slack-native
Your morning status report, on your terms.
Every weekday morning, a plain-English digest lands in your team channel: what broke overnight, what was filed, what is still open, and how much was recovered.
Questions
Frequently asked.
Quick answers about scope, control and how the credit-chaser actually behaves. Anything else — just email us.
Get started
One agent. One inbox. Your stack, finally watched.
We are onboarding a small first cohort of design partners. Tell us a little about your stack — Stripe, AWS, Datadog, GitHub — and we will reply within a business day.