I built an advanced n8n workflow designed to help a company process incoming emails, filter out irrelevant messages, and automatically generate high-quality draft replies using an AI agent.
The system is divided into four main stages:
Email Reading & Relevance Classification
An AI agent reads incoming emails and determines whether they are relevant client requests or irrelevant messages (spam, promotions, notifications, etc.).
Only relevant emails continue through the workflow.
Integrated Knowledge Base Access
A knowledge base is embedded inside the agent so it can rely on accurate, company-specific information (services, pricing, templates, policies, conversation history, etc.).
Automatic Draft Generation
A second AI agent generates clean, structured draft replies based on:
existing templates
pricing rules
team availability
the ongoing conversation thread
The workflow ensures no repetition and maintains consistent tone and quality.
Labeling & Delivery to the Company Inbox
The generated drafts are labeled, organized and uploaded back into the company’s email inbox.
The team can then review, adjust, and send the drafts instantly, saving significant time.
This workflow helps teams respond faster, maintain consistent communication quality, and reduce the time spent manually sorting and writing emails.
Technologies:
n8n, AI Agents, Knowledge Base Integration, Workflow Automation