n8n

n8n + Auggie Integration Guide

Automate AI-powered account research in your n8n workflows.

Overview

Auggie provides AI-powered account research that delivers composite scores, pain scores, talking points, and business problem analysis for any company. This guide shows you how to integrate Auggie into your n8n workflows.

This guide covers three approaches:

  • Webhook workflow (recommended) — fire-and-forget research with results delivered to a webhook
  • Polling workflow — submit research and poll until completion
  • Bulk research — research many companies at once from a spreadsheet

Prerequisites

  • 1
    Auggie API key — create one at /api-keys (starts with aug_)
  • 2
    n8n instance — cloud or self-hosted
  • 3
    Auggie webhook registered — for the webhook approach, see Webhook Workflow below

Single Research (Webhook)

Recommended
  1. 1
    Create a new workflow with a Webhook trigger node. Copy its Production URL.
  2. 2
    Register that URL with Auggie via POST /v1/webhooks/register.
  3. 3
    In a separate workflow (or Schedule trigger), add an HTTP Request node: POST to https://auggie.app/v1/research.
  4. 4
    Under Authentication, choose Generic Credential Type → Header Auth. Name: Authorization, Value: Bearer aug_your_key.
  5. 5
    Set body JSON (see below).
  6. 6
    When research completes, Auggie POSTs results to your Webhook trigger with scores, sections, and company_url.
  7. 7
    Add downstream nodes (Google Sheets, Slack, etc.).

Workflow Architecture

Workflow 1: Schedule / Form HTTP Request POST /v1/research
Workflow 2: Webhook Trigger Process Results Google Sheets / Slack / etc.

Request Body

{
  "company_url": "https://example.com"
}

Example Webhook Payload (received by your Webhook trigger)

{
  "job_id": "res_abc123",
  "company_url": "https://example.com",
  "status": "completed",
  "scores": {
    "composite": 82,
    "pain": 75,
    "product_fit": 88
  },
  "sections": {
    "talking_points": "They recently expanded into...",
    "business_problems": "Struggling with manual data...",
    "product_fit": "Strong alignment with automation..."
  }
}

Single Research (Polling)

If you prefer a single workflow without webhooks, you can poll for results.

  1. 1
    HTTP Request node: POST to https://auggie.app/v1/research with Header Auth (same as webhook approach).
  2. 2
    Body: {"company_url": "https://example.com"}
  3. 3
    Wait node: 5 seconds.
  4. 4
    Second HTTP Request: GET https://auggie.app/v1/research/{{ $('HTTP Request').item.json.job_id }} (same auth).
  5. 5
    IF node: {{ $json.status }} is not equal to completed.
  6. 6
    IF true → loop back to Wait node. IF false → continue to next action.

Workflow Flow

HTTP Request POST Wait 5s HTTP Request GET IF status != completed
↑ loops back to Wait  |  ↓ continues to next action

Bulk Research

Research many companies at once from a spreadsheet.

  1. 1
    Google Sheets node to read rows with a company_url column.
  2. 2
    HTTP Request POST to https://auggie.app/v1/research/bulk with body:
{
  "company_urls": {{ $json.rows.map(r => r.company_url) }},
  "name": "n8n bulk import"
}
3
Poll GET /v1/research/bulk/{id} for status, or use the webhook approach to receive results when complete.

Using Results

Once research is complete, the response contains scores and sections you can use in downstream nodes.

Key Fields

scores.composite

Overall research score

scores.pain

Pain signal strength

sections.talking_points

Ready-to-use talking points

sections.business_problems

Identified business problems

sections.product_fit

Product fit analysis

Example n8n Expressions

{{ $json.scores.pain }}

{{ $json.sections.talking_points }}

Troubleshooting

Issue Solution
401 Unauthorized Check API key starts with aug_, verify Header Auth config
Timeout / no response Research takes 30-60s, increase HTTP Request timeout to 120s
Webhook not firing Verify webhook is registered, check n8n webhook URL is accessible, workflow must be active
402 Payment Required No credits remaining, purchase at /billing/buy-credits
429 Rate Limited Reduce request frequency, check Retry-After header