One engine, end to end
GoRunner is what happens between an event and the work it should cause: trigger, pieces, AI, execution, and the record of all of it.
- 1
A trigger starts it
An email arrives, a webhook fires, a row changes, a schedule ticks. Triggers are push-based wherever the upstream allows it, so flows start in milliseconds, not on the next polling interval.
EmailWebhookScheduleRow change - 2
Pieces do the work
Each step is a piece: an action in Slack, Salesforce, Stripe, or any of 600+ others, or a logic step like branch, loop, and human approval. Data maps between steps visually, with the real payload in view.
Create caseBranchNotifyApprove - 3
AI fills the judgment gaps
Classification, extraction, drafting, summarizing: GoRunner AI steps handle the parts that used to need a person reading the thing. Or skip the canvas entirely and describe the whole flow in chat.
ClassifyDraft with ClaudeExtract JSON - 4
The engine runs it, fast
Flows execute on a compiled Go engine that schedules steps in-process instead of round-tripping a queue. Retries, timeouts, and backpressure are engine concerns, not your problem.
0.41s median run - 5
You see everything
Every run is recorded step by step with inputs, outputs, and timing. Replay runs, alert on failures, and answer 'what happened at 3am' with the actual payload.
ReplayTimelineAlerts
The whole operation, one screen
Flows, recent runs, health, and usage in a dashboard built for the person who gets asked "is the automation working?" The answer is a glance, and it is usually a green dot.
Runs today
41,209
Success rate
99.7%
p50 duration
0.38s
Go deeper
Ready when you say go.
Build your first flow in minutes. Free plan, no card, and the canvas is yours.
Free plan · 1,000 runs a month · self-host or cloud