Know your batch jobs are actually doing their job.
Truebatch quietly verifies every batch run and queue against the outcome you expected — and tells you the moment reality drifts from the plan.
Founding access · €29/mo locked for life · no spam, unsubscribe anytime.
row_count ≥ 5000 failed · paged #data-oncall · 6m ago.“Job succeeded” is not the same as “job worked.”
An exit code of 0 hides empty loads, stale data, broken schemas and stuck queues. Truebatch checks the outcome, not just the exit code.
Declarative selftests
Describe what a healthy run looks like in a tiny YAML manifest — row counts, freshness, schema, runtime budgets. Your expectations, versioned next to the job.
Continuous verification
Every run is checked the moment it finishes. Silent failures, half-loads and stuck queues surface in minutes — not when a customer emails you.
Alerts where you work
Slack, PagerDuty or email the instant a job drifts — deduplicated, with the failing check and the last good run attached. No more log-diving at 2am.
Your expectations, written down once — checked forever.
Define a healthy run in a small manifest your team reviews like code. Truebatch does the watching from there.
- Version-controlled checks. Selftests live in your repo and ship in pull requests.
- Outcome-based. Row counts, freshness, schema and SLAs — not just “did it run.”
- Zero agents. Read-only access, five-minute setup, nothing to maintain.
# nightly-billing-export pipeline: nightly-billing-export schedule: "0 2 * * *" checks: - row_count: min 5000 - freshness: max_age "3h" - runtime: max "15m" - schema: orders.v3 on_fail: - slack: #data-oncall - pagerduty: P1
Live in an afternoon.
Connect your pipeline
Point Truebatch at your scheduler or queue — cron, Airflow, Celery, Sidekiq, SQS. Read-only, five-minute setup, no agents to babysit.
Declare your selftests
Commit a selftest.yaml next to the job: expected rows, freshness windows, schema, SLAs. Reviewed in a pull request, like real code.
Get told first
We verify every run and ping you the moment something drifts — before your customers do. Mute, snooze, escalate, attach the last good run.
One simple price, locked for life.
Join the waitlist now and your founding rate never goes up — for every pipeline you add.
- Unlimited selftests & runs per pipeline
- Slack, PagerDuty & email alerting
- 90-day run history & audit trail
- Pull-request review for every check
- EU data residency, read-only access
Free during private beta. No card required to join the list.
Questions, answered.
Is this just another uptime monitor?
What can it watch?
When can I actually use it?
Where does my data live?
Stop finding failed jobs from an angry customer email.
Be first in line when Truebatch opens. Founding members lock the €29 price for life.
No spam. Just your invite and occasional build updates.