Durable execution
Workflows that survive deploys, restarts and network flakes. Guaranteed to finish, even if your API goes down at step seven.
v2.0 is live — multi-region queues shipped
The workflow engine your team deserves.
Mesh runs your team's operational workflows on infrastructure you already trust. Python SDK, webhooks, and a dashboard your ops team will actually use.
~/workflows/daily-reports.py
# Schedule a workflow with 3 lines. from mesh import workflow, task @workflow(schedule="0 9 * * *", region="eu-west") async def daily_reports(ctx): rows = await task("fetch_rows")(ctx.db) for row in rows: await task("email_report")(row) return {"sent": len(rows)}
Features
Everything you'd build yourself, already built.
Retries, observability, scheduling, backfills — the patterns every team ends up reimplementing, done once, properly.
Observable by default
Every step logged, every retry attributed, every failure reproducible. Built-in OpenTelemetry export.
Ops-friendly dashboard
Pause, resume and replay workflows from a web UI your non-engineers can safely use.
SDK, not framework
Python, TypeScript, Go. Runs in your code, your infrastructure, your repo. No lock-in.
Schedule anywhere
Cron, event-driven, webhook-triggered or called from another workflow. Multi-region out of the box.
Honest pricing
Per-workflow-run, not per-seat. Free below 50k runs/month. You'll never get a surprise invoice.
Ship your first workflow this afternoon.
Free during beta. No credit card. No sales call required — but we're here if you want one.