Get Started
Back to Templates

AI Chat Agent: Dumpling AI + GPT-4o to Auto-Save Local Business Data to Airtable

Created by

Created by: Yang || yang

Yang

Last update

Last update a month ago

Categories

Share

Who is this for?

This workflow is for digital marketers, small business owners, lead generation agencies, and VAs who need a scalable way to find and store local business leads using AI. It’s especially useful for teams that want to enrich leads with real-time news insights and save the structured data to Airtable.

What problem is this workflow solving?

Manually researching local businesses and staying up to date with relevant news is time-consuming and inefficient. This automation eliminates that burden by using Dumpling AI chat agents to generate leads and context, GPT-4o to summarize, and Airtable to store everything in one place.

What this workflow does

This AI workflow listens for a manual trigger in n8n and executes the following steps:

  1. Extracts local business leads using a Local Business Agent from Dumpling AI.
  2. Pulls current news related to the business type or location using a News Agent from Dumpling AI.
  3. Uses GPT-4o to combine both responses into a human-readable summary.
  4. Extracts structured lead data like name, category, and city.
  5. Saves the summary and lead data into Airtable for easy follow-up.

Setup

1. Create AI Agents in Dumpling AI

  • Sign in at Dumpling AI
  • Create two separate agents:
    • Local Business Agent: Designed to respond with structured lists of businesses by location and category.
    • News Agent: Designed to fetch relevant recent news and summaries about a specific industry or region.
  • After setting up each agent, copy the Agent Key from Dumpling AI. These keys will be required in the headers of your HTTP Request nodes in n8n.

2. Manual Trigger

  • This workflow begins with a manual trigger inside n8n, Which is the When chat message is recieved.
  • This makes it easy to test and reuse, especially during setup.

3. Get Local Business Data from Dumpling AI

  • The first HTTP Request node sends a prompt like List 5 top real estate companies in Atlanta with full address and services.
  • Include your Local Business Agent Key in the x-agent-key header.
  • The response will return a structured list of business leads.

4. Get News Context from Dumpling AI

  • The second HTTP Request node sends a prompt such as Give me the latest news related to the real estate market in Atlanta.
  • Use your News Agent Key in the header.
  • This fetches a brief set of recent news summaries relevant to the businesses being researched.

5. Use GPT-4o to Merge and Summarize

  • The GPT node combines the list of businesses and news into one coherent summary.
  • You can modify the prompt to output in paragraph format, bullet points, or structured notes.

6. Save Lead to Airtable

  • The Airtable node sends all structured fields into your selected base and table.
  • Be sure to connect your Airtable account and confirm the columns match exactly.

How to customize this workflow

  • Replace the prompt inside the HTTP node to focus on different types of businesses or cities.
  • Expand the GPT output to include additional lead info like websites, phone numbers, or emails if the agent includes them.
  • Add a webhook trigger to allow this flow to be run via a chatbot, external app, or button.
  • Link to HubSpot or another CRM to sync the leads automatically.
  • Duplicate the process to run for multiple industries in parallel.

Final Notes

  • You must create and configure your Dumpling AI agents first before running this workflow.
  • The Agent Keys from Dumpling AI are required in both HTTP Request nodes.
  • This flow is modular and flexible, ready for deeper CRM integrations.
  • The manual trigger is great for testing, but you can add a Webhook node to automate it.

This workflow helps you launch an intelligent lead gen process that combines location-targeted business discovery, AI-generated insights, and structured CRM-friendly output, all powered by Dumpling AI and OpenAI.