Step 2: Connect Apify and Configure Scraping Inputs

These nodes launch the LinkedIn job scraper and read the dataset results for processing.

1. Open Launch LinkedIn Scraper and set customBody to the JSON input provided, including the LinkedIn Jobs URL and "count": 100.
2. In Retrieve Dataset Records, set resource to Datasets, limit to 110, and datasetId to ={{ $json.defaultDatasetId }}.
3. Credential Required: Connect your Apify credentials in Launch LinkedIn Scraper and Retrieve Dataset Records.

Step 3: Filter and Normalize Company Records

This section filters to small companies, removes duplicates, excludes HR-related industries, and cleans company domains.

1. In Check Company Size, confirm the condition uses ={{ $json.companyEmployeesCount }} with operation lt and rightValue 250.
2. In Deduplicate Companies, set compare to selectedFields and fieldsToCompare to companyName.
3. In Exclude HR Industries, confirm the two notContains checks for Human Resources and Staffing and Recruiting use ={{ $json.industries }}.
4. Keep Finalize Company Info code as provided to extract the domain and wrap data under company.
5. In Verify Domain Presence, verify the exists check uses ={{ $json.company.domain }}.
6. Set Limit Company Batch to maxItems 5 to control the number of companies processed per run.

Limit Company Batch outputs to both Combine Data Streams and Apollo People Search in parallel.

Step 4: Fetch People and Email Data from Apollo

These nodes pull decision-makers and email addresses from Apollo, then prepare the data for merging.

1. In Apollo People Search, set url to https://api.apollo.io/api/v1/people/search and keep the jsonBody with {{ $json.company.domain }} inside q_organization_domains_list.
2. Credential Required: Connect your httpHeaderAuth credentials in Apollo People Search.
3. Keep the provided JavaScript in Clean Person Records to flatten and dedupe people while wrapping data under person.
4. In Apollo Email Lookup, set url to https://api.apollo.io/v1/people/match and jsonBody to ={"id":"{{ $json.person.personId }}"} .
5. Credential Required: Connect your httpHeaderAuth credentials in Apollo Email Lookup.
6. Keep the JavaScript in Clean Email Payload to wrap results under email.

Clean Person Records outputs to both Apollo Email Lookup and Combine Data Streams in parallel.

Step 5: Merge Company, Person, and Email Data

This stage merges three streams into a single lead profile for sheet insertion.

1. In Combine Data Streams, set numberInputs to 3 to merge company, person, and email data.
2. Keep the JavaScript in Assemble Lead Profile to map person.personId to email.id and append company fields.

Step 6: Connect Google Sheets and Append Leads

Lead profiles are added to your Google Sheet and then reloaded for email generation.

1. In Append Leads to Sheet, set operation to append, documentId to your sheet ID, and sheetName to gid=0.
2. Verify the column mappings use expressions like ={{ $json.email }}, ={{ $json.companyName }}, and ={{ $json.personLinkedInUrl }}.
3. Credential Required: Connect your Google Sheets credentials in Append Leads to Sheet.
4. In Retrieve Leads Sheet, select the same documentId and sheetName to read back the latest rows.
5. Credential Required: Connect your Google Sheets credentials in Retrieve Leads Sheet.

Append Leads to Sheet outputs to Retrieve Leads Sheet to begin the email enrichment flow.

Step 7: Validate Rows and Prepare Batches for AI

These checks ensure only valid leads without existing emails get processed and limit batch size.

1. In Validate Email & Content, confirm the conditions: ={{ $json.Subject }} is empty, ={{ $json["Email Body"] }} is empty, and ={{ $json.Email }} contains @.
2. In Batching Filter, keep the condition ={{ $json.row_number }} gt 83 to control processing batches.

Step 8: Configure AI Email Generation

The AI chain creates the subject and body using Gemini and enforces a structured response.

1. In AI Email Composer, keep the prompt text and variables such as {{ $json.Company }}, {{ $json['Job Title'] }}, and {{ $json.Occupation }}.
2. Open Gemini Chat Engine and connect it to AI Email Composer as the language model.
3. Open Structured Output Extractor and keep the jsonSchemaExample defining subject and email_body.
4. Credential Required: Connect your Google Gemini credentials in Gemini Chat Engine.

Structured Output Extractor is connected as the output parser for AI Email Composer—ensure credentials are added to Gemini Chat Engine, not the parser.

Step 9: Write Generated Emails Back to Google Sheets

The generated subject and email body are written back to the sheet using Email as the matching key.

1. In Write Email Back to Sheet, set operation to update and use matchingColumns Email.
2. Map columns using expressions: ={{ $('Batching Filter').item.json.Email }}, ={{ $json.output.subject }}, and ={{ $json.output.email_body }}.
3. Credential Required: Connect your Google Sheets credentials in Write Email Back to Sheet.

Step 10: Test and Activate Your Workflow

Run a full test to confirm scraping, enrichment, AI generation, and sheet updates work end-to-end.

1. Click Execute Workflow on Manual Start Trigger to run a live test.
2. Verify that Append Leads to Sheet adds new rows and Write Email Back to Sheet updates Subject and Email Body.
3. Check that Limit Company Batch processes only 5 companies and that parallel branches populate both Combine Data Streams and Apollo People Search.
4. When results look correct, toggle the workflow to Active for production use.