Purchase Request Approval

This use case shows how NeuronFlow handles a procurement workflow where request details arrive from a form, email, or chat intake, then move through policy-aware review before any purchase is approved.

Where This Fits

This workflow is a strong fit when:

• employees submit spend requests without a consistent format
• finance wants predictable approval rules by amount, department, or urgency
• managers need to request revisions instead of only saying yes or no
• the team wants a clear audit trail of who approved what and why

Typical Workflow Shape

1. A Webhook, form event, or inbox trigger starts the workflow.
2. A Set, Code, or AI Agent node normalizes the request into a consistent object.
3. AI or rules extract fields like amount, department, justification, vendor, and urgency.
4. A policy object is generated for Approval Group.
5. Approval Group pauses the workflow and routes the request to the right reviewers.
6. Reviewers can approve, reject, or request changes with comments.
7. If changes are requested, the workflow routes back to the requester with the reviewer notes attached.
8. Once approved, downstream nodes can notify finance, create records, or trigger the purchasing step.

Best Node Pattern

• Webhook or Trigger: start from a form submission, inbox intake, or internal request app
• Set, Code, or AI Agent: normalize the request and summarize vendor, amount, urgency, and justification
• Switch: separate low-value, standard requests from higher-risk or capex-style requests
• Set: create the approval policy object for reviewer routing
• Approval Group: route the decision to the correct manager, finance, or leadership reviewers
• Tool or HTTP Request: create the approved purchasing record in an ERP, procurement system, Airtable, or Google Sheets
• Tool: notify the requester or finance team in Slack, Gmail, or SMTP Email

Approval Policy Example

Use Approval Group policy mode when reviewer routing changes by context.

{
  "reviewers": [
    { "name": "Finance Manager", "email": "finance@example.com" },
    { "name": "Department Head", "email": "dept-head@example.com" }
  ],
  "review_mode": "all",
  "allow_comments": true,
  "allow_request_changes": true,
  "require_comment_on_reject": true,
  "require_comment_on_request_changes": true,
  "routing_metadata": {
    "severity": "high",
    "routing_reason": "Amount exceeds department threshold",
    "recommended_by": "ai"
  }
}

Why NeuronFlow Fits Well

• Policy mode keeps reviewer selection dynamic without losing deterministic control.
• Comments on approval, rejection, and request changes become structured workflow data.
• Rework loops preserve the original request context instead of starting over.
• Runtime metadata freezes the reviewer set and business context for auditability.

What Success Looks Like

The request moves through a consistent office workflow, approvers get the same structured brief every time, and finance can review the full decision path later without reconstructing it from email threads.