n8n Support Ticket RAG: Claude + Google Gemini Webhook
n8n Support Ticket RAG: Claude + Google Gemini Webhook
Regular price
£76.99
Regular price
£76.99
Sale price
Unit price
/
per
⬇
Instant Digital Download
∞
Unlimited Downloads
★
Lifetime Access in Your Account
Couldn't load pickup availability
🔥
128+ Sold
Popular with n8n builders
⚡
23 people viewing
High interest right now
✅
9 added today
Fast-moving digital product
n8n Support Ticket RAG: Claude + Google Gemini Webhook
Regular price
£76.99
Regular price
£76.99
Sale price
Unit price
/
per
Auto-Answer Support Tickets in n8n with RAG (Claude + Google Gemini)
Turn inbound support questions into faster, safer responses. This n8n Support Ticket RAG workflow receives inquiries via a header-authenticated webhook, masks personal data, retrieves similar past Q&A using Google Gemini embeddings, and uses Anthropic Claude to draft an auto-reply—only when confidence is high.
What this workflow does
- Receives customer inquiries through a webhook and normalizes common PII patterns using masking logic.
- Embeds and searches history: creates a Google Gemini embedding for the new question and queries an in-memory vector store for the most similar learned Q&A.
- Checks similarity confidence using a similarity threshold to decide whether an auto-reply is appropriate.
- Drafts replies with Anthropic Claude using only the retrieved reference Q&A, and blocks the auto-reply if Claude indicates it can’t answer or if the output is unsafe.
- Logs and learns: when auto-answered, it records the masked question, match score, and generated reply to an n8n Data Table and returns the reply to the webhook caller.
- Escalates when needed: if no safe auto-reply is possible, it logs the case as pending and returns a “waiting for staff” response.
- Updates the knowledge base: receives staff answers via a second header-authenticated webhook, applies regex + Claude masking, then adds the masked Q&A to the vector store and updates the case as learned (or marks it for review).
Use cases
- SaaS teams reducing time-to-first-response for common support issues.
- Automation engineers building an n8n RAG support assistant with auditable learning loops.
- Operations workflows where staff-approved answers continuously improve future auto-replies.
Technical details
- Webhook inputs (header-authenticated) for customer inquiries and staff answers
- PII masking with regex + Claude masking pass
- Google Gemini embeddings for retrieval
- In-memory vector store for similarity search
- Anthropic Claude for safe, reference-only reply drafting
- n8n Data Table logging for auto-answered, pending, and learned cases
