AI Agent Use Cases
Views —

AI Agent Use Cases

Added today

44All 189

Prompt to Prepare a Shopping Cart from a Recipe Photo

The author reports that Grok Bot receives a recipe photo, organizes the ingredients and quantities, and creates a shopping list by excluding ingredients already available at home.

Grok Bot
Life & Hobbies

1–20 of 189

Marketing & DesignWalkthrough

Filtering Leads and Sending Only Approved Emails

n8n filters out duplicate lead information and prepares contact emails for each customer. Only emails that have been approved by a person in Google Sheets are sent via Gmail.

n8n

How it works

  1. Customer information is received from Lead Intake and goes through Validation and Duplicate Check. This step filters out incorrect entries or leads that have already been received.
  2. After evaluating customers with AI Qualification and Lead Scoring, they are classified as HOT, WARM, or COLD. Then, the workflow connects to drafting a contact email for each customer.
  3. Customer information and approval status are recorded in Google Sheets. A manual approval step is placed before Gmail sending, so that drafts which have not been approved are not sent.

Results

This is a post introducing the first version of the workflow. Actual conversion rates or sending performance were not provided.

Notes

The original text discloses the processing order. The scoring formula, AI prompts, and node settings are not included, so it is not a fully reproducible tutorial. In the comments, there was a suggestion to record the reasoning behind the classification scores as well.

Source

n8n 커뮤니티 Original post ↗

Example prompt · n8n

Implement an n8n workflow that classifies incoming leads and sends only contact emails that have been approved by a person.

Input: Customer information form [fields], evaluation criteria [criteria table], product information [document], Google Sheets [document], sending account [connection]. Do not make up any scoring formulas not disclosed in the original text; first confirm the actual criteria and thresholds to be used.

After verifying required information and checking for duplicates, classify leads as HOT, WARM, or COLD. Record the basis for classification and any missing information, and draft an email for each customer. Do not invent customer circumstances or product benefits that are not in the materials.

In the sheet, record the customer identifier, classification, reasoning, draft, approval status, and sending result. Place an explicit approval check before the Gmail sending node, and if the draft changes after approval, require re-approval. Ensure that the same email is not sent multiple times on retries.

Deliverables: Workflow JSON, sheet template, and execution instructions. Test for missing information, duplicate leads, borderline scores, before and after approval, rejection, and sending failures to verify that no unapproved emails are sent.