Step 2: Connect Google Drive and Parse the Spreadsheet

After a new file appears, the workflow downloads it and parses the spreadsheet into rows.

1. In Drive File Download, set Operation to download and File ID to ={{ $json.id }}.
2. Credential Required: Connect your Google Drive credentials for Drive File Download (credentials are not configured).
3. In Spreadsheet Parse, enable Header Row so column names are recognized from the first row.
4. Connect Spreadsheet Parse to Iterate Records to process each row individually.

Step 3: Set Up AI Website Enrichment and Personalization

This section summarizes each company website and generates a personalized email opener based on whether a valid website was found.

1. In Web Summary Agent, set Text to =Please analyze the website at {{ $json.domain }} and create a structured business summary according to the template. Use the "Jina Markdown Fetcher" tool to get the website content. {{ $json.chatInput }}.
2. Anthropic Chat C is connected as the language model for Web Summary Agent — Credential Required: Connect your anthropicApi credentials to Anthropic Chat C.
3. Jina Markdown Fetcher is a tool sub-node for Web Summary Agent — Credential Required: Connect your httpHeaderAuth credentials to Jina Markdown Fetcher, and note that tool credentials are managed at the parent node level.
4. In OpenAI Website Check, enable JSON Output and keep the message template as provided; Credential Required: Connect your openAiApi credentials.
5. Branch Validation checks ={{ $json.message.content.hasWebsite }} and routes to No-Web Personalizer when false; otherwise it routes to Web Personalizer.
6. Anthropic Chat A powers No-Web Personalizer and Anthropic Chat B powers Web Personalizer — Credential Required: Connect your anthropicApi credentials to both nodes.
7. In No-Web Personalizer, keep the prompt and expressions such as {{ $('Iterate Records').item.json['First Name'] }} and {{ $('Iterate Records').item.json['Last Name'] }} intact.
8. In Web Personalizer, keep the summary reference {{ $('Web Summary Agent').item.json.output }} and the same name expressions from Iterate Records.

Execution Flow: Iterate Records outputs to both Finalize Batch and Web Summary Agent in parallel.

Step 4: Assign Personalization and Update the Sheet

Both personalization paths converge and the result is written back to the spreadsheet.

1. In Assign Personalization, set personalization to ={{ $json.output }} to capture the generated sentence.
2. In Update Sheet Rows, set Operation to update.
3. Set Document ID to ={{ $('Drive File Trigger').item.json.id }} and Sheet Name to ={{ $('Drive File Trigger').item.json.name }}.
4. Map columns to these expressions: domain → ={{ $('Spreadsheet Parse').item.json.domain }}, websiteSummary → ={{ $('Web Summary Agent').item.json.output }}, personalization → ={{ $json.personalization }}.
5. Credential Required: Connect your Google Sheets credentials to Update Sheet Rows (credentials are not configured).

Step 5: Send Completion Alert

After all rows are processed, the workflow builds a summary and notifies you in Telegram.

1. In Finalize Batch, review the JavaScript and replace the placeholders [YOUR_ID] and Lead Spreadsheet with your actual file ID and name.
2. In Telegram Ready Alert, set Chat ID to your Telegram chat and keep Text as =Lead Enrichment is ready: FileName: {{ $json.fileName }} fileURL: {{ $json.fileUrl }}.
3. Credential Required: Connect your telegramApi credentials to Telegram Ready Alert.

Step 6: Test and Activate Your Workflow

Run a manual test with a sample spreadsheet to confirm enrichment and messaging work end-to-end.

1. Click Execute Workflow and upload a test spreadsheet into the watched folder to trigger Drive File Trigger.
2. Confirm rows are parsed by Spreadsheet Parse and that Update Sheet Rows writes websiteSummary and personalization back to the sheet.
3. Verify a Telegram message is sent by Telegram Ready Alert with a valid file URL.
4. Once verified, toggle the workflow to Active to enable continuous processing for new files.