Skip to content
One platform. From customer conversation to completed work.Take a closer look
Documentation menu

Quickstart

One small process. A clear first result.

Start with a test conversation or a workflow. You do not need to configure the whole platform before making progress.

Product preview: workspace access opens at launch. Follow the launch.

Make it your workspace

  1. Choose the work you need.Compare Support, Workflows, and NodeHub, then choose a plan. Get started keeps that selection through account setup. Creating an account does not start a paid subscription.
  2. Choose your workspace.After signing in, choose an existing workspace or create one. Give a new workspace a name and optionally add your company website. Its favicon appears on the workspace card when available; an uploaded workspace logo takes precedence.
  3. Confirm the workspace and plan.A workspace owner or administrator reviews the selected plan in Plan & billing. Follow the checkout steps if subscriptions are available. A paid subscription is confirmed only after the payment provider verifies it.

Joining an existing team? Sign in with the email your team invited. If the workspace is missing, ask its owner to add that account, then reload NodeHub.

Support covers customer conversations. Workflows covers connected automation. NodeHub combines the two. Read Plans and billing for per-member prices, annual billing, and separate AI provider usage.

Start with Desk

  1. Open your workspace.Sign in to NodeHub using your workspace account. If this is your first visit, follow workspace setup above.
  2. Configure a support channel.An administrator opens Settings > Channels and completes the chosen email channel's configuration. Follow its verification instructions before sending a test message.
  3. Send a test request.Use an email address you control and a non-sensitive example. Confirm that the message appears in the correct business's conversation list.
  4. Assign, add context, and reply.Set an owner, add an internal note, and send a reply to your test address. Confirm the reply actually arrives and that an internal note stays inside the workspace.
  5. Check the next exchange.Reply from the test address and confirm that the conversation continues in the expected thread. Review the customer record and any follow-up task your process needs.

If the request does not arrive, check the channel setup and delivery status first. Saving channel settings is not the same as verifying inbound and outbound delivery.

Moving from n8n or Plain?

Use the import guide to preview an n8n workflow export or supported Plain support batches. Review the reported gaps, reconnect required accounts, and verify the results before switching your process.

Start with Automation

  1. Choose a small starting point.Open Workflows or choose a Hub workflow. A Hub install creates an unpublished draft in your workspace.
  2. Give related workflows a home.Optionally open Projects, choose New project, and name the process. Use a workflow's project selector to move it into that project. Moving a workflow organizes it without changing its trigger, published version, or access.
  3. Review the trigger and every step.Identify the input the workflow expects. Start with a manual test where the definition supports one, or use a dedicated test event for the configured trigger.
  4. Connect the required accounts.Configure each required connection and confirm verification. Select the intended account and operation in each step, then map its required inputs. Choose test records you are allowed to change.
  5. Add review before external writes.If the process updates another service, add a human approval where needed. Check the approved and rejected paths, including what happens when the request is refused.
  6. Test and inspect the result.Save the draft, inspect Preview, and use Run with controlled test input for the configured trigger. Read the run's step results. Confirm the actual provider outcome and check that a failure is visible.
  7. Publish when the behavior is right.Review connections, input mapping, trigger scope, and write actions before publishing. Publishing records a workflow version; configured automatic triggers can then start real work.

Testing a write action can affect the connected provider. Repeating a failed run also needs care: check whether the previous attempt already completed the external action.

When something needs attention

  • Sign-in is not configured: the server operator must configure authentication; repeatedly clicking Sign in will not fix it.
  • A control is missing: check your workspace membership and administrator permissions.
  • A connection is rejected: check the credential, provider scopes, and selected account.
  • A run is waiting: inspect the active step for an approval, event, or configured delay.
  • An AI step fails: verify its account, model availability, provider balance, tool permissions, and run limits.
Back to documentation