Step 2: Connect Twilio, Postgres, and Google Sheets

Attach credentials for the services used across the workflow so SMS, memory, and logging can run.

1. In Inbound SMS Trigger, Dispatch Intro SMS, Send Thank You SMS, Send Agent Reply SMS, and Notify Owner via SMS, add credentials: Credential Required: Connect your twilioApi credentials.
2. In Postgres Conversation Memory and Fetch Chat History, add credentials: Credential Required: Connect your postgres credentials.
3. In Append to Google Sheet, add credentials: Credential Required: Connect your googleSheetsOAuth2Api credentials.
4. For AI components, OpenAI Chat Model Alpha and OpenAI Chat Model Beta require: Credential Required: Connect your openAiApi credentials.

Step 3: Configure the Initial Outreach Flow

Send an introductory SMS shortly after the form is submitted.

1. In Pause Briefly, set Amount to 10 to add a short delay after form submission.
2. In Dispatch Intro SMS, set To to ={{ $json.Phone }}.
3. Set From in Dispatch Intro SMS to [YOUR_ID].
4. Set Message in Dispatch Intro SMS to Thanks for submitting the form on our website. Do you have a few minutes to answer a few questions?.

Step 4: Set Up the AI Qualification Loop

Route inbound SMS replies into the AI agent for qualification and decide whether to continue or complete the conversation.

1. Confirm the flow Inbound SMS Trigger → Hold for Reply → Lead Qualification Agent is connected as shown in the workflow.
2. In Postgres Conversation Memory, set Session Key to {{ $('Inbound SMS Trigger').item.json.data.from }} and Session ID Type to customKey. This memory node is attached to Lead Qualification Agent; ensure the Postgres credential is configured on Postgres Conversation Memory.
3. In OpenAI Chat Model Beta, set Model to gpt-4o and attach it to Lead Qualification Agent as the language model.
4. In Lead Qualification Agent, set Text to ={{ $json.data.body }} and keep Prompt Type as define.
5. Set System Message in Lead Qualification Agent to =You are a friendly real estate assistant helping qualify new leads for a realtor. Ask only three brief questions, then end the conversation politely. Keep the tone warm and professional. Your goal is to gather basic information to help the realtor follow up later. 1. Ask for their name. 2. Ask what type of property they’re looking for (e.g., house, condo, rental). 3. Ask what city or neighborhood they’re interested in. Once all three answers are collected, thank them and let them know a realtor will follow up soon. also add *** to the response.
6. In Conversation Completion Check, set the condition to String contains with Left Value ={{ $json.output }} and Right Value ***.
7. If the condition is false, the flow continues to Send Agent Reply SMS. Set To to ={{ $('Inbound SMS Trigger').item.json.data.from }} and Message to ={{ $json.output }}.

Step 5: Configure Completion Actions and Logging

When the agent signals completion, summarize the conversation and log it while sending final notifications.

1. Conversation Completion Check outputs to both Fetch Chat History and Send Thank You SMS in parallel.
2. In Send Thank You SMS, set To to ={{ $('Inbound SMS Trigger').item.json.data.from }} and Message to =Thanks, I'll give you a call soon..
3. In Fetch Chat History, set Operation to executeQuery and Query to SELECT * FROM n8n_chat_histories WHERE session_id = '{{ $('Inbound SMS Trigger').item.json.data.from }}'; .
4. In Aggregate History Records, set Aggregate to aggregateAllItemData.
5. In OpenAI Chat Model Alpha, set the model to gpt-4o-mini and attach it to Summarize Conversation as the language model.
6. In Structured Output Parser, keep JSON Schema Example as { "transcript": "transcript", "summary": "summary" } and connect it to Summarize Conversation as the output parser (no credentials needed for the parser itself).
7. In Summarize Conversation, set Text to ={{ $json.data }}, enable Has Output Parser, and set System Message to =output two fields a clean transcript of the conversation and a summary of the conversation.
8. In Append to Google Sheet, set Operation to append, Document ID to [YOUR_ID], and Sheet Name to gid=0 (Sheet1). Map columns: Phone = ={{ $('Inbound SMS Trigger').item.json.data.from }}, Summary = ={{ $json.output.summary }}, Transcript = ={{ $json.output.transcript }}.
9. In Notify Owner via SMS, set To to ={{ $('Inbound SMS Trigger').item.json.data.from }} and Message to =*** You just got a new lead. Here's a transcript of the interaction with our AI Bot. ***{{ $json.Summary }}***.

Step 6: Test and Activate Your Workflow

Validate the full SMS loop, AI responses, and logging before turning the workflow on.

1. Click Execute Workflow and submit a test entry via Form Submission Trigger with a real phone number.
2. Confirm you receive the intro SMS sent by Dispatch Intro SMS, then reply to trigger Inbound SMS Trigger and the AI loop through Lead Qualification Agent.
3. Verify that a non-completed conversation sends a reply from Send Agent Reply SMS, and that a completed conversation routes through the parallel path to Send Thank You SMS and Fetch Chat History.
4. Check that Append to Google Sheet creates a new row with the phone number, summary, and transcript, then confirm Notify Owner via SMS sends the final notification.
5. When everything works as expected, toggle the workflow to Active for production use.