This content is not available in your language yet.
Purpose
servEase Support uses AI and automation to speed up response times, but always with a human in the loop for anything customer-facing. This page explains how that works and what Support staff need to know to use it well.
RAG-Powered Responses
The Support Inbox (see Architecture & Infrastructure (Engineering space) → Support Inbox) can generate a draft response using the RAG Service, which retrieves relevant prior tickets and documentation to ground its answer instead of guessing.
Drafts are suggestions, not final answers — always read and verify before sending
If a draft is wrong or irrelevant, that's useful signal: check whether the underlying issue is missing from Known Issues / Troubleshooting by Module, since the RAG Service can only retrieve what's actually documented
n8n Workflows
n8n handles automation around the support flow, such as:
Routing/escalation triggers (e.g. flagging a ticket as needing L2 attention based on keywords or repeated contact)
_(document other active workflows here as they're built — e.g. auto-tagging, auto-closing resolved-and-confirmed tickets)_
Human Review Requirement
No AI-generated response should go to a customer unreviewed. This isn't just a quality safeguard — RAG answers are only as good as the documentation behind them, and an unreviewed wrong answer can create a bigger support problem than the original ticket.
Improving the System
The RAG Service gets better when the underlying documentation does. If you notice repeated bad AI suggestions on the same topic, that's a signal to update Known Issues or Troubleshooting by Module — not just a one-off annoyance to work around each time.