What this workflow does
This workflow allows you to ask natural language questions about your Airtable data using chat messages.
It turns those questions into real Airtable queries automatically.
The workflow then gets results from Airtable and sends back clear answers or even charts and maps.
This saves time and stops manual errors when searching data in Airtable.
Who should use this workflow
People managing many Airtable bases or tables with project information.
Users who want quick, accurate insights without writing formulas or searching manually.
Anyone who wants to talk to an assistant and get data answers fast.
Tools and services used
- n8n: Automates the workflow steps.
- OpenAI Chat Model: Understands user chat messages and guides Airtable queries.
- Airtable API: Searches bases and tables for requested data.
- Code nodes: Calculate aggregations like count, sum, and averages.
- HTTP Request nodes: Generate charts or map images from data.
How the workflow works
Inputs
User sends a normal chat message with a question or request about Airtable data.
The message contains the text and a session ID to keep conversation context.
Processing steps
- Chat capture: Webhook node receives the message and session info.
- AI understanding: OpenAI Chat Model processes text using a prompt designed to act like an Airtable assistant.
- Airtable search: Based on AI output, Airtable – Search records runs queries on bases and tables with filters.
- Error handling: If filters find no records, the workflow tries a broader search without filters.
- Calculations: Code node does math like counting records or averaging values when requested.
- Generating visuals: Another Code node or HTTP Request node creates graphs or map images for output.
- Combining results: Merge node merges all data for a full response.
Output
User gets simple answers, numbers, or charts back in chat or API reply.
Beginner step-by-step: How to build this in n8n
Import and open workflow
- Download the workflow file by clicking the Download button on this page.
- Go to your n8n editor dashboard.
- Click Import from File and select the downloaded workflow.
Configure credentials and settings
- Add your OpenAI API Key in the OpenAI Chat Model node.
- Add your Airtable API Key in the Airtable – Search records node.
- Update base IDs and table names in Airtable nodes to match your setup.
- Check the prompt text in the OpenAI Chat Model node – copy or change it if needed.
Test and activate
- Send a test chat message to the Webhook node to see if it processes correctly.
- Look at the outputs of each node to check for proper data flow and results.
- When satisfied, activate the workflow using the toggle switch in n8n.
- If self hosting n8n, make sure your server runs stable and API keys are safe. Learn more about self-host n8n.
Inputs and outputs explained
Input: Chat message text and session ID.
Output: Airtable records matching the query, aggregated numbers, and images for visual data.
Edge cases and errors
- If no records found with filter, workflow tries search without filters automatically.
- OpenAI API limits can cause delays; consider using smaller models or limiting requests.
- Session keys must be set properly to keep conversation context; otherwise AI loses track.
- Check Airtable base and table IDs carefully to avoid empty results.
Customization ideas
- Change AI model in OpenAI Chat Model node for different speed or accuracy.
- Add more math operations in the Code node like median or mode.
- Edit the system prompt to better match your Airtable fields and terms.
- Add more Airtable bases dynamically based on user input.
- Change the visual output to create other types of graphs or maps.
Summary of results
✓ Save hours by skipping manual Airtable searches.
✓ Get faster, clearer answers to data questions.
✓ Reduce errors from manual filtering or formulas.
✓ Receive charts and maps in responses for better understanding.