Step 2: Connect Data Tables

Configure the data table nodes that read lead rows and store enrichment plus email results.

1. In Retrieve Pending Rows, set Operation to get and verify the Data Table is email_linkedin_list. Confirm the filter condition: email_subject is isEmpty.
2. In Fetch Enrichment Rows, set Operation to get and use the same Data Table email_linkedin_list with the email_subject is isEmpty condition.
3. In Store Profile Data, set Operation to update, and map linkedin_profile_scrape to {{ $json.results.toJsonString() }} with the filter email = {{ $('Fetch Enrichment Rows').item.json.email }}.
4. In Store Company Data and Store Crunchbase Data, keep Operation as update and use the same email filter {{ $('Fetch Enrichment Rows').item.json.email }}.
5. In Write Email Results, set Operation to update and map email_body to {{ $('Assemble Email Context').item.json.email_content }} and email_subject to {{ $('Assemble Email Context').item.json.email_subject }}.

Step 3: Connect RapidAPI Enrichment Requests

Set the RapidAPI key and configure the parallel enrichment requests, delays, and result fetches.

1. In Set RapidAPI Token, set RapidAPI-Key to your key value, replacing [CONFIGURE_YOUR_API_KEY].
2. Confirm Fetch Enrichment Rows receives input from Set RapidAPI Token.
3. Fetch Enrichment Rows outputs to both LinkedIn Profile Request and LinkedIn Company Request and Crunchbase Request and Run Sub-Workflow (Configure Required) in parallel.
4. In LinkedIn Profile Request, set URL to https://cold-outreach-enrichment-scraper.p.rapidapi.com/company_url and use Query {{ $json.Linkedin_URL }} with header {{ $('Set RapidAPI Token').item.json['RapidAPI-Key'] }}.
5. In LinkedIn Company Request, set URL to https://cold-outreach-enrichment-scraper.p.rapidapi.com/company_url and use Query {{ $json.company_linkedin }} with header {{ $('Set RapidAPI Token').item.json['RapidAPI-Key'] }}.
6. In Crunchbase Request, set URL to https://cold-outreach-enrichment-scraper.p.rapidapi.com/url_search and use Query {{ $json.Crunchbase_URL }} with header {{ $('Set RapidAPI Token').item.json['RapidAPI-Key'] }}.
7. In Delay Profile Fetch, Delay Company Fetch, and Delay Crunchbase Fetch, keep Unit set to minutes and adjust the actual delay duration as needed.
8. In Profile Fetch Results, Company Fetch Results, and Crunchbase Fetch Results, set URL to https://cold-outreach-enrichment-scraper.p.rapidapi.com/results and set task_id to {{ $('LinkedIn Profile Request').item.json.task_id }}, {{ $('LinkedIn Company Request').item.json.task_id }}, and {{ $('Crunchbase Request').item.json.task_id }} respectively.
9. In Run Sub-Workflow (Configure Required), select the sub-workflow in Workflow and decide whether to keep Wait For Sub-Workflow as false.

Step 4: Set Up AI Email Generation

Configure the AI agent and structured parsers to generate personalized outreach content.

1. In Personalized Outreach Writer, keep Prompt Type as define and ensure the prompt includes the enrichment placeholders like {{ $json.First_name }}, {{ $json.Company_Name }}, and {{ $json.linkedin_profile_scrape }}.
2. Attach Structured Parse Schema to Personalized Outreach Writer so the agent outputs JSON with email_subject and email_content.
3. In Assemble Email Context, map fields to expressions: email_subject = {{ $json.output.email_subject }}, email_content = {{ $json.output.email_content }}, and email = {{ $('Batch Iteration').item.json.email }}.
4. Connect Gemini Chat Engine as the language model for Personalized Outreach Writer.

Step 5: Configure Review Routing and Batch Flow

Review emails, approve or revise, and manage looping through batches.

1. In Quality Review Agent, keep the defined prompt and ensure it receives {{ $json.email_subject }} and {{ $json.email_content }} as inputs.
2. Attach Approval Schema Parser to Quality Review Agent to enforce structured responses.
3. In Approval Decision, set the condition to check Left Value {{ $json.output }} contains APPROVED.
4. Confirm the flow: Quality Review Agent → Approval Decision; approved emails go to Write Email Results while non-approved output returns to Retrieve Pending Rows for reprocessing.
5. Ensure Batch Iteration receives from Retrieve Pending Rows and loops back from Write Email Results for the next row.

Step 6: Configure Output Updates

Write final email subject and body back to the data table once approved.

1. In Write Email Results, ensure email_body maps to {{ $('Assemble Email Context').item.json.email_content }} and email_subject maps to {{ $('Assemble Email Context').item.json.email_subject }}.
2. Confirm the update filter uses email = {{ $('Batch Iteration').item.json.email }} so only the correct row is updated.

Step 7: Test and Activate Your Workflow

Run a controlled test to validate enrichment, AI output, and data table updates.

1. Click Execute Workflow and start with Manual Launch Trigger to run a test on a single row.
2. Verify that enrichment data is stored in Store Profile Data, Store Company Data, and Store Crunchbase Data, then confirm AI output in Write Email Results.
3. Successful execution shows new values in email_subject and email_body within the email_linkedin_list table.
4. Once validated, activate the workflow and trigger it from the parent workflow using Triggered by Parent Flow.