Step 2: Connect Google Sheets

These nodes pull meeting records and write research and follow-up outputs back to your sheet.

1. Open Review Call Records and set Document to Mock Data and Sheet to Meeting Data.
2. In Review Call Records, confirm the filter for company_overview with Lookup Value set to FILL to fetch rows needing research.
3. Open Write Research Update, set Operation to update, and confirm matching on Email.
4. Open Write Follow-up Outputs, set Operation to update, and confirm matching on Email.
5. Credential Required: Connect your googleSheetsOAuth2Api credentials on Review Call Records, Retrieve Testimonials, Get Product Catalog, Write Research Update, and Write Follow-up Outputs.

Step 3: Set Up Research Orchestration Agent

This agent researches each prospect and writes structured insights back to the sheet.

1. Open Research Orchestration Agent and set Text to =Please research these sales call attendees Name: {{ $json.Name }} Email: {{ $json.Email }} Company Name: {{ $json['Company Name'] }} Website: {{ $json.Website }} Business Type: {{ $json['Business Type'] }} Project: {{ $json.Project }}.
2. Ensure the agent uses OpenAI Chat Engine as its language model.
3. Attach Structured JSON Extractor as the output parser to enforce the six-field JSON schema.
4. Attach tools Tavily Company Lookup, Get Product Catalog, and Write Research Update to the agent.
5. Credential Required: Connect your openAiApi credentials on OpenAI Chat Engine.

Step 4: Set Up Follow-up Copy Assistant

This agent turns research results into a personalized subject line, email, and SMS.

1. Open Follow-up Copy Assistant and set Text to =Name: {{ $('Review Call Records').item.json.Name }} Email: {{ $('Review Call Records').item.json.Email }} Company Name: {{ $('Review Call Records').item.json['Company Name'] }} Website: {{ $('Review Call Records').item.json.Website }} Business Type: {{ $('Review Call Records').item.json['Business Type'] }} Project: {{ $('Review Call Records').item.json.Project }} company_overview: {{ $json.output.company_overview }} tech_stack: {{ $json.output.tech_stack }} company_updates: {{ $json.output.company_updates }} primary_solution: {{ $json.output.primary_solution }} solution_2: {{ $json.output.solution_2 }} solution_3: {{ $json.output.solution_3 }}.
2. Ensure the agent uses OpenAI Chat Engine 2 as its language model.
3. Attach Message Output Parser as the output parser with the subject/email/text_message schema.
4. Attach tools Retrieve Testimonials and Write Follow-up Outputs to log the finalized content.
5. Credential Required: Connect your openAiApi credentials on OpenAI Chat Engine 2.

Step 5: Confirm Output Mappings

Verify that both update tools write the correct AI fields back to the sheet using matching email addresses.

1. In Write Research Update, verify the column mappings use expressions like ={{ $json.Email }} and ={{ /*n8n-auto-generated-fromAI-override*/ $fromAI('company_overview', ``, 'string') }}.
2. In Write Follow-up Outputs, confirm mappings for email_subject, email_text, and sms use $fromAI(...) expressions and Email matches ={{ $('Review Call Records').item.json.Email }}.
3. Make sure Write Research Update and Write Follow-up Outputs both target Meeting Data in Mock Data.

Step 6: Test and Activate Your Workflow

Run a manual test to confirm research output, follow-up copy, and Google Sheets updates are working end-to-end.

1. Click Execute Workflow from Manual Run Starter.
2. Confirm that Review Call Records pulls rows where company_overview is FILL.
3. Verify Write Research Update fills company_overview, tech_stack, company_updates, and solution fields.
4. Verify Write Follow-up Outputs fills email_subject, email_text, and sms for the same email address.
5. When successful, toggle the workflow to Active for production use.