💬 Verify AI draft answers with Pearl Hybrid Intelligence and OpenAI

⚡ 135 views · 💬 Support Chatbots

Description

Pearl Hybrid Intelligence: AI draft + expert verification (across 100+ expert domains)

Deliver higher-confidence answers by combining an AI assistant with professional human verification.

Pearl provides access to a network of 20,000 licensed experts (including JDs, MDs, PhDs, DVMs, CPAs, engineers) across 100+ domains - Legal and Health are just examples, and the prompts can be tuned for many other areas.

What this template does

• Lets users ask questions in a conversational format (your app sends the conversation history).
• Collects any missing intake details before answering.
• Generates a clear draft answer using AI.
• Sends the draft to a Pearl expert for professional verification.
• Returns the verified answer back to the user, including who verified it.

How it works (non-technical)

1. A user asks a question.
2. If the assistant needs more context, it asks one focused follow-up question.
3. Once the assistant has enough information, it drafts a complete answer.
4. That draft is automatically reviewed by a qualified Pearl expert.
5. The user receives the verified answer, along with expert attribution.

Setup steps (10–15 minutes)

• Connect your OpenAI credential in n8n (used for intake + drafting).
• Add your Pearl MCP Server API key (used for expert verification). You can request a demo access key here: https://www.pearl.com/enterprise/contact-get-started
• Activate the workflow and send a test request using the example payload below.

Customization

• The intake and draft answer prompts can be adjusted to match end-user needs (domain, tone, risk policy, compliance requirements).
• You can also modify the response format (fields returned, disclaimers, attribution formatting) without changing the core flow.

Input payload format (conversation history)

Send a JSON body like:

{
  "model": "gpt-4o-mini",
  "messages": [
    { "role": "user", "content": "My question..." }
  ]
}

## 🔗 Nodes Used

HTTP Request, Webhook, MCP Client

## 📥 Import

Download [`workflow.json`](workflow.json) and import into n8n:
**Workflow menu → Import from File**

[📖 Importing guide](../../../docs/importing-templates.md) · [🔑 Credential setup](../../../docs/credential-setup.md)