Start with what happens.
Choose a supported trigger: a webhook, schedule, ticket event, or manual run. Map the incoming data into the steps that need it.
Connect events, tools, AI decisions, and people in a workflow you can inspect. Start from a webhook, a schedule, or a customer request.
Build automation around the job. Bring in the services it needs, define the next step, and see what happened when it runs.
Choose a supported trigger: a webhook, schedule, ticket event, or manual run. Map the incoming data into the steps that need it.
Use conditions to choose a path, a bounded AI decision for judgment, and a human approval where the next action deserves review.
Inspect run history and step outputs. Find the failed operation, review its configuration, and decide how the process should recover.
Choose a model from your configured AI accounts, connect the knowledge and tools it needs, and put approval steps before actions that require a person.
A customer message. A scheduled check. An event in another app. Give each one a clear next step.
Bring your model, knowledge, and selected operations into the same workflow.
Ask for approval before the next action. Give your team an explicit place in the process.
Conditions and AI decisions route each request to the next appropriate action.
Inspect workflow runs and step results. Follow the work back to what actually happened.
Use supported service actions, HTTP requests, and agent tools to connect the work.
Bring customer conversations into the same workspace as the process. Connect the request, the decision, and the follow-up without rebuilding the story in another app.
Explore NodeHub DeskYes. A workflow can begin with a webhook, schedule, or manual run. Customer conversations are another possible starting point, not a prerequisite for building automation.
No. Use explicit conditions and service actions when the process is predictable. Add AI decisions or agents where their configured capabilities fit the job.
Model usage depends on the AI account and provider you configure. Connecting a model does not make its usage free. Check the provider's charges and set run limits before enabling an AI workflow.
Run history records step results and errors. Inspect the failure and use the controls supported by that step. Configure retries carefully for external writes so a repeated request does not create a duplicate action.
No. Installing a Hub workflow creates a private draft. Review its configuration, connect the required accounts, and test the process before publishing it.
Start with one process. Connect the tools that make it work.
Get started with Automation