Every lead read, qualified and routed before anyone opens the inbox.
Connect the mailbox, the CRM and a model. Let a workflow classify, enrich and assign, and let a human approve the ones that matter.
Three workflows teams build here first.
- 01
Inbound inbox triage: classify every email, extract company and intent, create or update the CRM record.
- 02
Lead routing by territory, size and product line, with the assignee notified in chat and the reply drafted.
- 03
Renewal and upsell nudges from CRM stage changes, with a human sign-off before anything customer-facing goes out.
What runs, step by step.
Each new email is classified; leads are enriched, created in the CRM and handed to an owner, and a reply is drafted once a person approves it.
- 01New email
- 02Classify intent
- 03Lead or not?
- 04Enrich company
- 05Create CRM contact
- 06Tell the assignee
- 07Approve the reply
- 08Draft the reply
Every step is a node type in the catalog. Diagram only: no data, no live run.
Before you start
- A Gmail connection for the shared inbox (or an Outlook one).
- A HubSpot or Salesforce connection that can create contacts.
- A credential for the AI provider that does the classifying.
- Your routing rules (territory, size, product line) written as branch conditions, and a named approver for replies.
- A Clearbit account if you want enrichment, and a Slack channel or DM for the assignee.
What you get
Every inbound email is classified; leads land in the CRM with an owner who has been told, and no reply leaves until a person has approved it.
Limitations
- Classification is only as good as your categories and the model. Review the first runs before you trust the branch.
- The Anthropic integration has no nodes of its own: its models are used through the AI nodes with an Anthropic credential.
- Model calls count against the workspace AI budget. When the budget is reached, the AI steps stop until it resets or you raise it.
- Enrichment needs your own Clearbit account; OCNORA does not supply the data.
Sales inbox: qualify and route
Poll the sales mailbox, classify each email as lead / existing customer / vendor, create Iris records for new leads and forward the rest to the right team.
See each step's output, and approve the risky ones.
Illustrated with a support workflow; the run view and the approval request look the same for every workflow.
Built for this, not adapted to it.
Gmail, HubSpot and Salesforce nodes
First-class nodes for the mailbox and the CRM, plus Outlook, Pipedrive and Zoho. OAuth handled by the connection, not by the workflow.
A lead-router template to start from
The sales inbox template fetches mail, classifies it, branches and routes. Import it, swap the connections, run it.
AI classify and extract as nodes
Structured classification and field extraction with the provider of your choice. Outputs are typed, so the branch after them is a plain condition.
Human in the loop
A human prompt node pauses the run for approval with a timeout and a fallback. Big-ticket replies wait for a person; the rest go out.
AI spend you can see
Every model call lands in a cost ledger with a per-workspace budget. Sales automation cannot quietly become your largest AI bill.
Business reports per workflow
Hours saved, money saved and reliability per workflow, so the sales ops lead can show the number without a spreadsheet.
Start from one of these.
Each one imports into a workspace; swap the connections and run it.
- Sales inbox: qualify and routePoll the sales mailbox, classify each email as lead / existing customer / vendor, create Iris records for new leads and forward the rest to the right team.
- Inbox triage: classify and route emailEvery 15 minutes, fetch the next unread email, classify it (urgent / billing / general) and route urgent ones to the on-call inbox with an auto-acknowledgement.
- AI screen, human approve, then sendClassify an inbound request, ask a reviewer to approve or reject it, and email the requester the outcome.
Apps this kind of workflow usually touches.
Each one has its own page with operations, triggers and auth.
Gmail
HubSpot
Salesforce
Anthropic
SlackRead before the first live run.
- How to test a workflow before it touches anything realStatic validation, a dry run against real config with side effects held back, then a first live run on one record. A practical order for shipping automations safely.
- How to keep an AI agent’s spending close to the number you setA retry loop, a huge tool result or a growing prompt can turn one agent into a runaway bill. Here’s how per-tenant budgets and a circuit breaker limit it.
- How much a dry run actually proves before you trust it liveA dry run executes your workflow with side effects suppressed and shows each step’s output shape. That shows the path resolves, not that the result is right.
See a workflow run on your own use case.
Book a technical demo. Bring an n8n file or pick a template, and we build it with you on the canvas while Workflow Health checks it before it ever runs.